Data-Over-Cable Service Interface Specifications
DCA - MHAv2
Remote PHY Specification
CM-SP-R-PHY-I01-150615
ISSUED
Notice
This DOCSIS® specification is the result of a cooperative effort
undertaken at the direction of Cable Television Laboratories, Inc. for the
benefit of the cable industry and its customers. You may download, copy,
distribute, and reference the documents herein only for the purpose of
developing products or services in accordance with such documents, and
educational use. Except as granted by CableLabs® in a separate written
license agreement, no license is granted to modify the documents herein
(except via the Engineering Change process), or to use, copy, modify or
distribute the documents for any other purpose.
This document may contain references to other documents not owned or
controlled by CableLabs. Use and understanding of this document may
require access to such other documents. Designing, manufacturing,
distributing, using, selling, or servicing products, or providing services,
based on this document may require intellectual property licenses from
third parties for technology referenced in this document. To the extent
this document contains or refers to documents of third parties, you agree
to abide by the terms of any licenses associated with such third-party
documents, including open source licenses, if any.
 Cable Television Laboratories, Inc. 2014-2015
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
DISCLAIMER
This document is furnished on an "AS IS" basis and neither CableLabs nor its members provides any representation or
warranty, express or implied, regarding the accuracy, completeness, noninfringement, or fitness for a particular
purpose of this document, or any document referenced herein. Any use or reliance on the information or opinion in this
document is at the risk of the user, and CableLabs and its members shall not be liable for any damage or injury
incurred by any person arising out of the completeness, accuracy, or utility of any information or opinion contained in
the document.
CableLabs reserves the right to revise this document for any reason including, but not limited to, changes in laws,
regulations, or standards promulgated by various entities, technology advances, or changes in equipment design,
manufacturing techniques, or operating procedures described, or referred to, herein.
This document is not to be construed to suggest that any company modify or change any of its products or procedures,
nor does this document represent a commitment by CableLabs or any of its members to purchase any product whether
or not it meets the characteristics described in the document. Unless granted in a separate written agreement from
CableLabs, nothing contained herein shall be construed to confer any license or right to any intellectual property. This
document is not to be construed as an endorsement of any product or company or as the adoption or promulgation of
any guidelines, standards, or recommendations.
2
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
Document Status Sheet
Document Control Number
CM-SP-R-PHY-I01-150615
Document Title
Remote PHY Specification
Revision History
I01 - Released 06/15/2015
Date:
Status:
Distribution Restrictions:
June 15, 2015
Work in
Progress
Draft
Author
Only
CL/Member
Issued
Closed
CL/ Member/
Vendor
Public
Key to Document Status Codes
Work in Progress
An incomplete document designed to guide discussion and generate feedback, and may
include several alternative solutions for consideration.
Draft
A document in Specification format considered largely complete, but lacking review by
Members and Technology Suppliers. Drafts are susceptible to substantial change during the
review process.
Issued
A generally public document that has undergone Member and Technology Supplier review,
cross-vendor interoperability, and is available for Certification testing. Issued
Specifications are subject to the Engineering Change (EC) Process.
Closed
A static document, reviewed, tested, validated, and closed to further ECs.
Trademarks
CableLabs® is a registered trademark of Cable Television Laboratories, Inc. Other CableLabs marks are listed at
http://www.cablelabs.com/certqual/trademarks. All other marks are the property of their respective owners.
06/15/15
CableLabs®
3
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
Table of Contents
1
SCOPE ..................................................................................................................................................................9
1.1
1.2
1.3
2
Introduction and Purpose ...............................................................................................................................9
MHAv2 Interface Documents ........................................................................................................................9
Requirements and Conventions .....................................................................................................................9
REFERENCES .................................................................................................................................................. 10
2.1
2.2
2.3
Normative References.................................................................................................................................. 10
Informative References ................................................................................................................................ 12
Reference Acquisition.................................................................................................................................. 12
3
TERMS AND DEFINITIONS .......................................................................................................................... 13
4
ABBREVIATIONS AND ACRONYMS .......................................................................................................... 17
5
TECHNICAL OVERVIEW ............................................................................................................................. 22
5.1
Introduction ................................................................................................................................................. 22
5.2
System Diagram........................................................................................................................................... 23
5.2.1
Hub Access Network ............................................................................................................................ 23
5.2.2
Optical Access Network ....................................................................................................................... 23
5.2.3
Coax Access Network........................................................................................................................... 24
5.2.4
Location of the Remote PHY Device .................................................................................................... 24
5.2.5
Location of the CCAP-Core ................................................................................................................. 24
5.3
System Architecture..................................................................................................................................... 24
5.3.1
System Components ............................................................................................................................. 24
5.4
Remote PHY Device Architecture ............................................................................................................... 26
5.5
Remote PHY Operation ............................................................................................................................... 26
5.5.1
R-DEPI and R-UEPI............................................................................................................................ 26
5.5.2
Remote DTI .......................................................................................................................................... 27
5.6
Latency ........................................................................................................................................................ 27
5.7
MHAv2 Summary ....................................................................................................................................... 28
6
RPD INITIALIZATION ................................................................................................................................... 29
6.1
Overview ..................................................................................................................................................... 29
6.2
Security ........................................................................................................................................................ 30
6.3
Network Authentication ............................................................................................................................... 30
6.3.1
Problem Definition .............................................................................................................................. 30
6.3.2
Authentication from an Untrusted Portion of the Network .................................................................. 30
6.3.3
802.1x Authentication .......................................................................................................................... 31
6.4
Address Assignment .................................................................................................................................... 36
6.4.1
DHCP Options ..................................................................................................................................... 37
6.4.2
Failures ................................................................................................................................................ 38
6.4.3
Security Implications ........................................................................................................................... 39
6.5
Time of Day ................................................................................................................................................. 39
6.5.1
ToD Acquisition ................................................................................................................................... 39
6.5.2
ToD Conflicts and Problems................................................................................................................ 39
6.5.3
ToD Security Implications ................................................................................................................... 39
6.6
Connection to CCAP-Cores ......................................................................................................................... 39
6.6.1
Core Types ........................................................................................................................................... 39
6.6.2
Connection Process ............................................................................................................................. 40
6.6.3
Connection to Auxiliary Cores............................................................................................................. 44
6.7
Synchronization ........................................................................................................................................... 46
6.7.1
Synchronization Failures ..................................................................................................................... 46
6.8
Connectivity................................................................................................................................................. 47
4
CableLabs®
06/15/15
Remote PHY Specification
7
CM-SP-R-PHY-I01-150615
SECURE SOFTWARE DOWNLOAD ............................................................................................................ 48
7.1
Introduction ................................................................................................................................................. 48
7.2
Overview ..................................................................................................................................................... 48
7.3
RPD Software Upgrade Procedure .............................................................................................................. 50
7.4
Software Code Upgrade Requirements ........................................................................................................ 52
7.4.1
Code File Processing Requirements .................................................................................................... 52
7.4.2
Code File Access Controls ................................................................................................................... 52
7.4.3
RPD Code Upgrade Initialization ....................................................................................................... 53
7.4.4
Code Signing Guidelines ..................................................................................................................... 55
7.4.5
Code Verification Requirements .......................................................................................................... 55
7.4.6
DOCSIS Interoperability ..................................................................................................................... 56
7.4.7
Error Codes ......................................................................................................................................... 56
7.5
Security Considerations (Informative) ......................................................................................................... 58
8
X.509 CERTIFICATE PROFILE AND MANAGEMENT ........................................................................... 59
8.1
Certificate Management Architecture Overview ......................................................................................... 59
8.2
RPD Certificate Storage and Management in the RPD ............................................................................... 59
8.3
Certificate Processing and Management in the CCAP-Core ....................................................................... 59
8.3.1
CCAP-Core Certificate Management Model ....................................................................................... 60
8.3.2
Certificate Validation .......................................................................................................................... 60
8.4
Certificate Revocation ................................................................................................................................. 61
8.4.1
Certificate Revocation Lists ................................................................................................................. 61
8.4.2
Online Certificate Status Protocol ....................................................................................................... 62
9
PHYSICAL PROTECTION OF KEYS IN THE RPD................................................................................... 64
10
SYSTEM OPERATION (NORMATIVE) ................................................................................................... 65
10.1 DOCSIS Upstream Scheduling .................................................................................................................... 65
10.1.1
Centralized Scheduling Requirements ................................................................................................. 65
10.2 Daisy-chaining of the Backhaul Ethernet Port ............................................................................................. 65
10.2.1
Backhaul Daisy-chaining Requirements .............................................................................................. 66
10.3 Networking Considerations ......................................................................................................................... 66
10.3.1
Per Hop Behavior ................................................................................................................................ 66
10.3.2
DiffServ Code Point Usage .................................................................................................................. 67
10.3.3
Packet Sequencing ............................................................................................................................... 67
10.3.4
Network MTU ...................................................................................................................................... 68
10.4 Pilot Tone Generation .................................................................................................................................. 68
11
MULTIPLE CCAP-CORE OPERATION .................................................................................................. 70
11.1
11.2
11.3
11.4
11.5
Introduction ................................................................................................................................................. 70
RPD Startup with Multiple Cores ................................................................................................................ 70
Downstream Channel Constraint Table ....................................................................................................... 71
Resource Sets and Auxiliary Resource Assignment .................................................................................... 72
RPD Reads and Writes ................................................................................................................................ 73
ANNEX A
A.1
A.2
A.3
DEPI MTU (NORMATIVE) ............................................................................................................. 74
L2TPv3 Lower Layer Payload Size ............................................................................................................. 74
Maximum Frame Size for DEPI .................................................................................................................. 74
Path MTU Discovery ................................................................................................................................... 74
ANNEX B
GCP USAGE (NORMATIVE) ......................................................................................................... 76
B.1
RPD Configuration with GCP(UCD) .......................................................................................................... 76
B.2
RPD Upstream Scheduler with GCP(DSx) .................................................................................................. 76
B.3
R-PHY Control Protocol.............................................................................................................................. 77
B.3.1
RCP over GCP EDS Message.............................................................................................................. 77
B.3.2
RCP over GCP EDS Response Messages ............................................................................................ 77
06/15/15
CableLabs®
5
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
B.3.3
B.3.4
B.3.5
B.3.6
B.3.7
B.3.8
B.3.9
B.3.10
B.3.11
B.3.12
B.3.13
B.3.14
RCP over GCP Device Management Message .................................................................................... 78
RCP over GCP Notify Message ........................................................................................................... 79
RCP TLV Format ................................................................................................................................. 79
RCP Message Structure ....................................................................................................................... 79
RCP Messages Types ........................................................................................................................... 80
RCP Protocol Rules ............................................................................................................................. 80
Extensibility ......................................................................................................................................... 80
Protocol Versioning ............................................................................................................................. 80
Information Model Extensibility .......................................................................................................... 80
Vendor Specific Extensions .................................................................................................................. 81
Inclusion of DOCSIS Messages ........................................................................................................... 81
RCP Message Examples ...................................................................................................................... 81
ANNEX C
R-DEPI EXTENSIONS OF R-UEPI USAGE (NORMATIVE) .................................................... 84
ANNEX D
MPEG STREAM ANALYSIS (NORMATIVE) ............................................................................. 85
ANNEX E
CERTIFICATE HIEARCHY AND PROFILES (NORMATIVE)................................................ 86
E.1
E.2
E.3
E.4
E.5
E.6
CableLabs Root CA Certificate ................................................................................................................... 87
CableLabs Device CA Certificate ................................................................................................................ 88
RPD Certificate ............................................................................................................................................ 89
CableLabs Service Provider CA Certificate ................................................................................................ 90
CCAP-Core Device Certificate .................................................................................................................... 91
AAA Server Certificate ............................................................................................................................... 92
APPENDIX I
I.1
I.2
I.3
PLANT SWEEP IN A DISTRIBUTED ARCHITECTURE (INFORMATIVE) ..................... 93
Plant Sweep Using Transmitter and Receiver Capabilities .......................................................................... 93
Hardware Module in the Node..................................................................................................................... 93
R-PHY Node API Support ........................................................................................................................... 93
APPENDIX II
ACKNOWLEDGEMENTS ...................................................................................................... 95
List of Figures
Figure 1 - Logical View of RPD Internals ................................................................................................................... 22
Figure 2 - Remote PHY System Diagram ................................................................................................................... 23
Figure 3 - MHAv2 Reference Architecture for DOCSIS Signaling and Provisioning ................................................ 25
Figure 4 - Remote PHY Device Block Diagram ......................................................................................................... 26
Figure 5 - R-PHY Internal Components ...................................................................................................................... 26
Figure 6 - RPD Initialization ....................................................................................................................................... 29
Figure 7 - Remote PHY: Trusted Domain and Untrusted Domain .............................................................................. 30
Figure 8 - Authentication Network Diagram ............................................................................................................... 31
Figure 9 - Network Authentication Signaling.............................................................................................................. 32
Figure 10 - RPD Topologies for 802.1x ...................................................................................................................... 33
Figure 11 - RPD Authentication using 802.1x............................................................................................................. 35
Figure 12 - DHCP Network Diagram .......................................................................................................................... 36
Figure 13 - DHCP Signaling........................................................................................................................................ 37
Figure 14 - CCAP-Cores DHCP Suboption IPv4 ........................................................................................................ 38
Figure 15 - CCAP-Cores DHCP Suboption IPv6 ........................................................................................................ 38
Figure 16 - Process for Connecting to the Principal Core ........................................................................................... 43
6
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
Figure 17 - Process for Connecting to Auxiliary Cores ............................................................................................... 45
Figure 18 - Typical Code Validation Hierarchy .......................................................................................................... 50
Figure 19 - CRL Framework ....................................................................................................................................... 61
Figure 20 - OCSP Framework ..................................................................................................................................... 62
Figure 21 - Certificate Hiearchy .................................................................................................................................. 86
List of Tables
Table 1 - PHBs and Recommended DSCP Values ...................................................................................................... 67
Table 2 - MTU of DEPI (for PSP) ............................................................................................................................... 74
Table 3 - GCP Encoding for the Upstream Scheduler ................................................................................................. 76
Table 4 - RCP Encodings for GCP EDS Messages ..................................................................................................... 77
Table 5 - RCP Encodings for GCP EDS Normal Response Messages ........................................................................ 77
Table 6 - RCP Encodings for GCP EDS Error Response Messages............................................................................ 78
Table 7 - RCP Encodings for GCP Device Management Messages ............................................................................ 78
Table 8 - RCP Encodings for GCP Notify Messages .................................................................................................. 79
Table 9 - Summary of RCP Messages ......................................................................................................................... 80
Table 10 - CableLabs Root CA Certificate .................................................................................................................. 87
Table 11 - CableLabs Device CA Certificate .............................................................................................................. 88
Table 12 - RPD Certificate .......................................................................................................................................... 89
Table 13 - CableLabs Service Provider CA Certificate ............................................................................................... 90
Table 14 - CCAP-Core Device Certificate .................................................................................................................. 91
Table 15 - AAA Server Certificate .............................................................................................................................. 92
06/15/15
CableLabs®
7
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
This page intentionally left blank.
8
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
1 SCOPE
1.1
Introduction and Purpose
Modular Headend Architecture version 2 (MHAv2)/Remote PHY technology allows a CMTS to support an IP-based
digital HFC plant. In an IP-based digital HFC plant, the fiber portion utilizes a baseband network transmission
technology such as Ethernet, EPON (Ethernet over Passive Optical Networks), GPON (Gigabit Passive Optical
Network), or any Layer 2 technology that would support a fiber-based Layer 1. MHAv2 uses a Layer 3 pseudowire
between a CCAP-Core and a series of Remote PHY devices. One of the common locations for a Remote PHY device
at an optical node device that is located at the junction of the fiber and coax plants.
1.2
MHAv2 Interface Documents
A list of the documents in the MHAv2 family of specifications is provided below. For updates, refer to
http://www.cablelabs.com/specs/specification-search/.
Designation
Title
CM-SP-R-PHY
Remote PHY Specification
CM-SP-R-DEPI
Remote Downstream External PHY Interface Specification
CM-SP-R-UEPI
Remote Upstream External PHY Interface Specification
CM-SP-GCP
Generic Control Plane Specification
CM-SP-R-DTI
Remote DOCSIS Timing Interface Specification
CM-SP-R-OOB
Remote Out-of-Band Specification
CM-SP-R-OSSI
Remote PHY OSS Interface Specification
NOTE: MHAv2 does not explicitly use any of the original Modular Headend Architecture specifications.
1.3
Requirements and Conventions
In this specification, the following convention applies any time a bit field is displayed in a figure. The bit field should
be interpreted by reading the figure from left to right, then from top to bottom, with the MSB being the first bit to read
and the LSB being the last bit to read.
Throughout this document, the words that are used to define the significance of particular requirements are
capitalized. These words are:
"MUST"
This word means that the item is an absolute requirement of this specification.
"MUST NOT"
This phrase means that the item is an absolute prohibition of this specification
"SHOULD"
This word means that there may exist valid reasons in particular circumstances to ignore
this item, but the full implications should be understood and the case carefully weighed
before choosing a different course.
"SHOULD NOT" This phrase means that there may exist valid reasons in particular circumstances when
the listed behavior is acceptable or even useful, but the full implications should be
understood and the case carefully weighed before implementing any behavior described
with this label.
"MAY"
06/15/15
This word means that this item is truly optional. One vendor may choose to include the
item because a particular marketplace requires it or because it enhances the product, for
example; another vendor may omit the same item.
CableLabs®
9
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
2 REFERENCES
At the time of publication, the editions indicated were valid. All references are subject to revision, and users of this
document are encouraged to investigate the possibility of applying the most recent editions of the documents listed
below. References are either specific (identified by date of publication, edition number, version number, etc.) or
non-specific. For a nonspecific reference, the latest version applies.
2.1
Normative References
In order to claim compliance with this specification, it is necessary to conform to the following standards and other
works as indicated, in addition to the other requirements of this specification. Notwithstanding, intellectual property
rights may be required to use or implement such normative references.
[CANN]
CableLabs’ Assigned Names and Numbers, CL-SP-CANN-I13-150515, May 15, 2015, Cable
Television Laboratories, Inc.
[CCAP-OSSI v3.1]
DOCSIS 3.1 CCAP OSSI Specification, CM-SP-CCAP-OSSIv3.1-I04-150611, June 11,
2015, Cable Television Laboratories, Inc.
[CM-OSSI v3.1]
DOCSIS 3.1 Cable Modem OSSI Specification, CM-SP-CM-OSSIv3.1-I04-150611. June 11,
2015, Cable Television Laboratories, Inc.
[DEPI]
Downstream External PHY Interface Specification, CM-SP-DEPI-I08-100611, June 11,
2010, Cable Television Laboratories, Inc.
[DRFI]
DOCSIS Downstream Radio Frequency Interface, CM-SP-DRFI-I14-131120, November 20,
2013, Cable Television Laboratories, Inc.
[FIPS 140-2]
Federal Information Processing Standards Publication (FIPS PUB) 140-2, Security
Requirements for Cryptographic Modules, May 2001.
[FIPS 180-4]
Federal Information Processing Standards Publication (FIPS PUB) 180-4, Secure Hash
Standard, May 2014.
[GCP]
Generic Control Plane Specification, CM-SP-GCP-I01-150615, June 15, 2015, Cable
Television Laboratories, Inc
[IANA-PORTS]
IANA, Port Numbers, June 2004.
[IEEE 802.1ae]
IEEE Std 802.1ae-2006, IEEE Standard for Local and metropolitan area networks—Media
Access Control (MAC) Security, August 2006.
[IEEE 802.1q]
IEEE Std 802.1q-2003, Virtual Bridged Local Area Networks, May 2003.
[IEEE 802.1x]
IEEE Std 802.1x-2010, IEEE Standard for Local and metropolitan area networks--Port-Based
Network Access Control, February 2010.
[IEEE 802.3]
IEEE Std 802.3™-2002, Part 3: Carrier Sense Multiple Access With Collision Detection
(CSMA/CD) Access Method and Physical Layer Specifications, March 2002.
[IEEE 1588]
IEEE Std 1588-2008, IEEE Standard for a Precision Clock Synchronization Protocol for
Networked Measurement and Control Systems, July 2008.
[ISO 13818-1]
ISO/IEC 13818-1:2013, Information Technology - Generic Coding of Moving Pictures and
Associated Audio Information. Part 1: System, May 23, 2013.
[ITU-T J.83]
ITU-T Recommendation J.83 (4/97), Digital multi-programme systems for television sound
and data services for cable distribution.
[MULPI v3.0]
DOCSIS MAC and Upper Layer Protocols Interface Specification,
CM-SP-MULPIv3.0-I27-150528, March 5, 2015, Cable Television Laboratories, Inc.
[MULPI v3.1]
DOCSIS MAC and Upper Layer Protocols Interface Specification,
CM-SP-MULPIv3.1-I06-150611, June 11, 2015, Cable Television Laboratories, Inc.
[PHY v3.0]
DOCSIS 3.0 Physical Layer Specification, CM-SP-PHYv3.0-I26-150305, March 5, 2015,
Cable Television Laboratories, Inc.
10
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
[PHY v3.1]
DOCSIS 3.1 Physical Layer Specification, CM-SP-PHYv3.1-I06-150611, June 11, 2015,
Cable Television Laboratories, Inc.
[SECv3.0]
DOCSIS 3.0 Security Specification, CM-SP-SECv3.0-I15-130808, August 8, 2013, Cable
Television Laboratories, Inc.
[SECv3.1]
DOCSIS 3.1 Security Specification, CM-SP-SECv3.1-I03-150611, June 11, 2015, Cable
Television Laboratories, Inc.
[PKCS#7]
RSA Laboratories, PKCS #7: Cryptographic Message Syntax Standard, An RSA Laboratories
Technical Note, Version 1.5, Revised November 1, 1993.
[R-DEPI]
Remote Downstream External PHY Interface Specification, CM-SP-R-DEPI-I01-150615,
June 15, 2015, Cable Television Laboratories, Inc.
[R-DTI]
Remote DOCSIS Timing Interface Specification, CM-SP-R-DTI-I01-150615, June 15, 2015,
Cable Television Laboratories, Inc.
[R-OOB]
Remote Out-of-Band Specification, CM-SP-R-OOB-I01-150615, June 15, 2015, Cable
Television Laboratories, Inc.
[R-OSSI]
Remote PHY OSS Interface Specification, CM-SP-R-OSSI-D02-150408, April 8, 2015,
Cable Television Laboratories, Inc.
[R-UEPI]
Remote Upstream External PHY Interface Specification, CM-SP-R-UEPI-I01-150615, June
15, 2015, Cable Television Laboratories, Inc.
[RFC 768]
IETF RFC 768, User Datagram Protocol, August 1980.
[RFC 791]
IETF RFC 791, Internet Protocol-DARPA, September 1981.
[RFC 868]
IETF RFC 768, Time Protocol, May 1983.
[RFC 1191]
IETF RFC 1191, MTU Path Discovery, November 1990.
[RFC 1350]
IETF RFC 1350, The TFTP Protocol (Revision 2), July 1992.
[RFC 1945]
IETF RFC 1945, Hypertext Transfer Protocol -- HTTP/1.0, May 1996.
[RFC 1981]
[RFC 2131]
IETF RFC 1981, Path MTU Discovery for IP version 6, August 1996.
IETF RFC 2131, Dynamic Host Configuration Protocol, March 1997.
[RFC 2348]
IETF RFC 2348, TFTP Blocksize Option, May 1998.
[RFC 2597]
IETF RFC 2597, Assured Forwarding PHB Group, June 1999.
[RFC 2616]
IETF RFC 2616, Hypertext Transfer Protocol -- HTTP/1.1, June 1999.
[RFC 2983]
IETF RFC 2983, Differentiated Services and Tunnels, October 2000.
[RFC 3246]
IETF RFC 3246, An Expedited Forwarding PHB (Per-Hop Behavior), March 2002.
[RFC 3260]
IETF RFC 3260, New Terminology and Clarifications for Diffserv, April 2002.
[RFC 3308]
IETF RFC 3308, Layer Two Tunneling Protocol (L2TP) Differentiated Services Extension,
November 2002.
[RFC 3748]
IETF RFC 3748, Extensible Authentication Protocol (EAP)
[RFC 3931]
IETF RFC 3931, Layer Two Tunneling Protocol - Version 3 (L2TPv3), March 2005.
[RFC 4131]
IETF RFC 4131, Management Information Base for Data Over Cable Service Interface
Specification (DOCSIS) Cable Modems and Cable Modem Termination Systems for
Baseline Privacy Plus, September 2005.
[RFC 4307]
IETF RFC 4307, Cryptographic Algorithms for Use in the Internet Key Exchange Version 2
(IKEv2), December 2005.
[RFC 5216]
IETF RFC 5216, IEAP-TLS Authentication Protocol, March 2008.
[RFC 5247]
IETF RFC 5247, EAP Key Management Framework, August 2008.
[RFC 5280]
IETF RFC 5280, Internet X.509 Public Key Infrastructure Certificate and Certificate
Revocation List (CRL) Profile, May 2008.
06/15/15
CableLabs®
11
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
[RFC 6960]
IETF RFC 6960, I.509 Internet Public Key Infrastructure Online Certificate Status Protocol OCSP, June 2013.
[RFC 7296]
IETF RFC 7296, Internet Key Exchange Protocol Version 2 (IKEv2), October 2014.
[RSA 1]
RSA Laboratories, PKCS #1: RSA Encryption Standard. Version 1.5, RSA Security, Inc.,
Bedford, MA, November 1993.
RSA Laboratories, PKCS #3: Diffie-Hellman Key Agreement Standard, Version 1.4, RSA
Security, Inc., Bedford, MA, November 1993.
Refers to RFC 3232 “Assigned Number” by the IETF, Jan 2002. This spec refers to the IANA
web page which is http://www.iana.org/assignments/enterprise-numbers.
ITU-T Recommendation X.509: Information Technology - Open Systems Interconnection –
The Directory: Public-key and Attribute Certificate Frameworks, March 2000.
[RSA 3]
[Vendor ID]
[X.509]
2.2
Informative References
This document uses the following informative references:
[IANA-L2TP]
IANA, Layer Two Tunneling Protocol (L2TP) Parameters.
[ISO 8802-2]
ISO/IEC 8802-2: 1994 (IEEE Std 802.2: 1994) - Information technology –
Telecommunications and information exchange between systems - Local and
metropolitan area networks - Specific requirements - Part 2: Logical link control.
[RFC 3140]
IETF RFC 3140, Per Hop Behavior Identification Codes, June 2001.
2.3
Reference Acquisition
•
Cable Television Laboratories, Inc., 858 Coal Creek Circle, Louisville, CO 80027;
Phone +1-303-661-9100; Fax +1-303-661-9199; http://www.cablelabs.com
•
Federal Information Processing Standards: 100 Bureau Drive, Mail Stop 3200, Gaithersburg, MD 20899-3200.
Phone +1-301-975-4054; Fax +1-301-926-8091. http://csrc.nist.gov/publications/fips/.
•
The Institute of Electrical and Electronics Engineers, Inc., Internet: http://standards.ieee.org
•
International Organization for Standardization (ISO), Tel.: +41 22 749 02 22, Fax: +41 22 749 01 55,
www.standardsinfo.net
•
Internet Assigned Numbers Authority, IANA, Internet: http://www.iana.org
•
Internet Engineering Task Force (IETF) Secretariat, 48377 Fremont Blvd., Suite 117, Fremont, California 94538,
USA, Phone: +1-510-492-4080, Fax: +1-510-492-4001. http://www.ietf.org.
•
ITU Recommendations: Place des Nations, CH-1211, Geneva 20, Switzerland. Phone +41-22-730-51-11; Fax
+41-22-733-7256. http://www.itu.int.
•
Public Key Cryptography Standards: RSA Security Inc. 174 Middlesex Turnpike, Bedford, MA 01730. Phone
+1-781-515-5000; Fax 781-515-5010. http://www.rsasecurity.com/rsalabs/.
•
SCTE, Society of Cable Telecommunications Engineers, 140 Philips Road, Exton, PA 19341-1318,
Phone+1-800-542-5040; Fax+1-610-363-5898. http://www.scte.org/default.aspx/.
12
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
3 TERMS AND DEFINITIONS
This specification uses the following terms:
Bonded Channels
A logical channel comprising multiple individual channels.
Cable Modem (CM)
A modulator-demodulator at subscriber locations intended for use in conveying data
communications on a cable television system.
CCAP-Core
A CCAP device that uses MHAv2 protocols to interconnect to an RPD.
The network (generally gigabit Ethernet) that connects a CCAP-Core to an RPD.
Converged
Interconnect Network
Customer Premises
Equipment (CPE)
Equipment at the end user's premises; may be provided by the service provider.
Data Rate
Throughput, data transmitted in units of time usually in bits per second (bps).
Decibels (dB)
Ratio of two power levels expressed mathematically as dB = 10log10(POUT/PIN).
Decibel-Millivolt
(dBmV)
Unit of RF power expressed in decibels relative to 1 millivolt, where dBmV = 20log10(value in
mV/1 mV).
Downstream (DS)
•
Transmissions from CMTS to CM. This includes transmission from the CCAP-Core to the
EQAM, as well as the RF transmissions from the EQAM to the CM
•
RF spectrum used to transmit signals from a cable operator's headend or hub site to
subscriber locations.
Dynamic Host
Configuration
Protocol (DHCP)
A network protocol enabling a server to automatically assign an IP address to a network
element.
Edge QAM Modulator A headend or hub device that receives packets of digital video or data. It re-packetizes the video
or data into an MPEG transport stream and digitally modulates the digital transport stream onto
(EQAM)
a downstream RF carrier using quadrature amplitude modulation (QAM).
Flow
A stream of packets in DEPI used to transport data of a certain priority from the CCAP-Core to
a particular QAM channel of the EQAM. In PSP operation, there can exist several flows per
QAM channel.
Gbps
Gigabits per second
Gigahertz (GHz)
A unit of frequency; 1,000,000,000 or 109 Hz
GigE (GE)
Gigabit Ethernet (1 Gbps)
Hertz (Hz)
A unit of frequency; formerly cycles per second
Hybrid Fiber/Coax
(HFC) System
A broadband bidirectional shared-media transmission system using optical fiber trunks
between the headend and the fiber nodes, and coaxial cable distribution from the fiber nodes to
the customer locations.
Institute of Electrical
and Electronic
Engineers (IEEE)
A voluntary organization which, among other things, sponsors standards committees and is
accredited by the American National Standards Institute (ANSI).
Internet Engineering
Task Force (IETF)
A body responsible for, among other things, developing standards used in the Internet.
Internet Protocol (IP) An Internet network-layer protocol
kilohertz (kHz)
Unit of frequency; 1,000 or 103 Hz; formerly kilocycles per second
L2SS
Layer 2 Specific Sublayer. DEPI is an L2SS of L2TPv3.
06/15/15
CableLabs®
13
CM-SP-R-PHY-I01-150615
L2TP Access
Concentrator (LAC)
Data-Over-Cable Service Interface Specifications
If an L2TP Control Connection Endpoint (LCCE) is being used to cross-connect an L2TP
session directly to a data link, we refer to it as an L2TP Access Concentrator (LAC). An LCCE
may act as both an L2TP Network Server (LNS) for some sessions and an LAC for others, so
these terms must only be used within the context of a given set of sessions unless the LCCE is,
in fact, single purpose for a given topology.
L2TP Attribute Value The L2TP variable-length concatenation of a unique Attribute (represented by an integer), a
length field, and a Value containing the actual value identified by the attribute.
Pair (AVP)
L2TP Control
Connection
An L2TP control connection is a reliable control channel that is used to establish, maintain, and
release individual L2TP sessions, as well as the control connection itself.
L2TP Control
Connection Endpoint
(LCCE)
An L2TP node that exists at either end of an L2TP control connection. May also be referred to
as an LAC or LNS, depending on whether tunneled frames are processed at the data link (LAC)
or network layer (LNS).
L2TP Control
Connection ID
The Control Connection ID field contains the identifier for the control connection, a 32-bit
value. The Assigned Control Connection ID AVP, Attribute Type 61, contains the ID being
assigned to this control connection by the sender. The Control Connection ID specified in the
AVP must be included in the Control Connection ID field of all control packets sent to the peer
for the lifetime of the control connection. Because a Control Connection ID value of 0 is used
in this special manner, the zero value must not be sent as an Assigned Control Connection ID
value.
L2TP Control
Message
An L2TP message used by the control connection.
L2TP Data Message
Message used by the data channel.
L2TP Endpoint
A node that acts as one side of an L2TP tunnel.
L2TP Network Server If a given L2TP session is terminated at the L2TP node and the encapsulated network layer
(L3) packet processed on a virtual interface, we refer to this L2TP node as an L2TP Network
(LNS)
Server (LNS). A given LCCE may act as both an LNS for some sessions and an LAC for others,
so these terms must only be used within the context of a given set of sessions unless the LCCE
is in fact single purpose for a given topology.
L2TP Pseudowire
(PW)
An emulated circuit as it traverses a packet-switched network. There is one Pseudowire per
L2TP Session.
L2TP Pseudowire
Type
The payload type being carried within an L2TP session. Examples include PPP, Ethernet, and
Frame Relay.
L2TP Session
An L2TP session is the entity that is created between two LCCEs in order to exchange
parameters for and maintain an emulated L2 connection. Multiple sessions may be associated
with a single Control Connection.
L2TP Session ID
A 32-bit field containing a non-zero identifier for a session. L2TP sessions are named by
identifiers that have local significance only. That is, the same logical session will be given
different Session IDs by each end of the control connection for the life of the session. When the
L2TP control connection is used for session establishment, session IDs are selected and
exchanged as Local Session ID AVPs during the creation of a session. The Session ID alone
provides the necessary context for all further packet processing, including the presence, size,
and value of the Cookie, the type of L2-Specific Sublayer, and the type of payload being
tunneled.
MAC Domain
A grouping of Layer 2 devices that can communicate with each other without using bridging or
routing. In DOCSIS, it is the group of CMs that are using upstream and downstream channels
linked together through a MAC forwarding entity.
Maximum
Transmission Unit
(MTU)
Maximum size of the Layer 3 payload of a Layer 2 frame.
Mbps
Megabits per second
14
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
Media Access Control Used to refer to the Layer 2 element of the system which would include DOCSIS framing and
signaling.
(MAC)
Modulation Error
Ratio (MER)
A unit of frequency; 1,000,000 or 106 Hz
The ratio of average signal constellation power to average constellation error power – that is,
digital complex baseband signal-to-noise ratio – expressed in decibels.
Microsecond (µs)
10-6 second
Millisecond (ms)
10-3 second
Modulation Error
Ratio (MER)
The ratio of the average symbol power to average error power.
Multiple System
Operator (MSO)
A corporate entity that owns and/or operates more than one cable system.
Nanosecond (ns)
10-9 second
Packet Identifier
(PID)
PID (system): A unique integer value used to identify elementary streams of a program in a
single or multi-program Transport Stream as described in section 2.4.3 of ITU-T Rec. H.222.0
[ISO 13818-1]
Physical Media
Dependent (PMD)
Sublayer
A sublayer of the Physical layer which is concerned with transmitting bits or groups of bits over
particular types of transmission link between open systems and which entails electrical,
mechanical, and handshaking procedures.
Pilot tones
Required in the HFC network to ensure that amplifiers in the network are operating correctly.
Amplifiers use these tones to adjust gain and keep signals at the appropriate output level.
Precision Time
Protocol
A protocol used to synchronize clocks throughout a network.
Program Clock
Reference (PCR)
A timestamp in the Video Transport Stream from which decoder timing is derived.
Pseudowire
An IP tunnel between two points in an IP network.
QAM channel (QAM
ch)
Analog RF channel that uses quadrature amplitude modulation (QAM) to convey information
Quadrature
Amplitude
Modulation (QAM)
A modulation technique in which an analog signal’s amplitude and phase vary to convey
information, such as digital data.
Megahertz (MHz)
Radio Frequency (RF) In cable television systems, this refers to electromagnetic signals in the range 5–1000 MHz.
Radio Frequency
Interface
Term encompassing the downstream and the upstream radio frequency interfaces.
Request For
Comments (RFC)
A technical policy document of the IETF; these documents can be accessed on the World Wide
Web at http://www.rfc-editor.org/.
Request-Grant Delay
Time
The time from when a CM requests bandwidth, using an uncontended bandwidth request
(REQ), to when it receives a MAP message with the granted transmit opportunity in it.
Remote-PHY Device
The Remote-PHY Device (RPD) is a device in the network which implements the
Remote-PHY specification to provide conversion from digital Ethernet transport to analog RF
transport.
Session
An L2TP data plane connection from the CCAP-Core to the QAM channel. There must be one
session per QAM Channel. There is one DEPI pseudowire type per session. There may be one
MPT flow or one or more PSP flows per session. Multiple sessions may be bound to a single
control connection.
StopCCN
L2TPv3 Stop-Control-Connection-Notification message
Trivial File Transfer
Protocol (TFTP)
A file transfer protocol. Generally used for automated transfer of configuration or boot files
between machines
06/15/15
CableLabs®
15
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
Upconverter
A device used to change the frequency range of an analog signal, usually converting from a
local oscillator frequency to an RF transmission frequency.
Upstream (US)
•
Transmissions from CM to CMTS. This includes transmission from the EQAM to
CCAP-Core as well as the RF transmissions from the CM to the EQAM.
•
RF spectrum used to transmit signals from a subscriber location to a cable operator’s
headend or hub site.
Upstream Channel
Descriptor (UCD)
The MAC Management Message used to communicate the characteristics of the upstream
physical layer to the cable modems.
Video on Demand
(VoD) System
System that enables individuals to select and watch video content over a network through an
interactive television system.
16
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
4 ABBREVIATIONS AND ACRONYMS
This specification uses the following abbreviations:
ACK
L2TPv3 Explicit Acknowledgement message
ADC
Analog-to-Digital Converter
AF
Assured Forwarding
AGC
Automatic Gain Control
API
Application Programming Interface
ATM
Asynchronous Transfer Mode
AVP
L2TPv3 Attribute Value Pair
BPI
Baseline Privacy Interface
CA
Certificate Authority
CAK
Connectivity Association Key
CCAP™
Converged Cable Access Platform
CDN
L2TPv3 Call-Disconnect-Notify message
CIN
Converged Interconnect Network
CLI
Command Line Interface
CM
Cable Modem
CMCI
Cable Modem to Customer Premises Equipment Interface
CMTS
Cable Modem Termination System
CPE
Customer Premises Equipment
CRC
Cyclic Redundancy Check
CRC16
CRC of length 16
CRL
Certificate Revocation List
CSMA
Carrier Sense Multiple Access
CVC
Code Verification Certificate
CVS
Code Verification Signature
CW
Continuous Wave
DAC
Digital-to-Analog Converter
dB
Decibels
dBmV
Decibel-Millivolt
DCA
Distributed CCAP Architecture
DEPI
Downstream External PHY Interface
DER
Distinguished Encoding Rules
DF
Don’t Fragment (bit)
DHCP
Dynamic Host Configuration Protocol
DHCPv4
Dynamic Host Configuration Protocol version 4
DHCPv6
Dynamic Host Configuration Protocol version 6
DOCSIS
Data-Over-Cable Service Interface Specifications
DOCSIS-MPT
(D-MPT)
DOCSIS MPT Mode
DPI
SCTE-35/Digital Program Insertion
06/15/15
CableLabs®
17
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
DRFI
Downstream Radio Frequency Interface
DS
Downstream
DSA
Dynamic Service Flow Add
DSCP
Differentiated Services Code Point
DSC
Dynamic Service Flow Change
DSD
Dynamic Service Flow Delete
DTA
Digital Television Adapter
DTI
DOCSIS Timing Interface
DTS
DOCSIS Timestamp, 32-bit
EAP
Extensible Authentication Protocol
EAPoL
Extensible Authentication Protocol over LAN
EBIF
Enhanced TV Binary Interchange Format
EF
Expedited Forwarding
EQAM
Edge QAM
ERM
Edge Resource Manager
ERMI
Edge Resource Manager Interface
ETSI
European Telecommunications Standards Institute
FDM
Frequency Division Multiplex
FQDN
Fully Qualified Domain Name
Gbps
Gigabits per second
GCP
Generic Control Plane
GE
Gigabit Ethernet (Gig E))
GHz
Gigahertz
HDLC
High-Level Data Link Control
HELLO
L2TPv3 Hello message
HFC
Hybrid Fiber/Coax
HMAC
Hash-based Message Authentication Code
Hz
Hertz
I-CCAP
Integrated CCAP
ICCN
L2TPv3 Incoming-Call-Connected message
ICMP
Internet Control Message Protocol
I-CMTS
Integrated CMTS
ICRP
L2TPv3 Incoming-Call-Reply message
ICRQ
L2TPv3 Incoming-Call-Request message
ID
Identifier
IEEE
Institute of Electrical and Electronic Engineers
IETF
Internet Engineering Task Force
IKE
Internet Key Exchange
IP
Internet Protocol
IPv4
Internet Protocol version 4
IPv6
Internet Protocol version 6
18
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
ISO
International Standards Organization
ITU
International Telecommunications Union
ITU-T
Telecommunication Standardization Sector of the International Telecommunication Union
kbps
Kilobits per second
kHz
Kilohertz
L2SS
Layer 2 Specific Sublayer
L2TP
Layer 2 Transport Protocol
L2TPv3
Layer 2 Transport Protocol version 3
L3
Layer 3
LAC
L2TP Access Concentrator
LCCE
L2TP Control Connection Endpoint
LNS
L2TP Network Server
LSB
Least Significant Bit
MAC
Media Access Control
MAP
Upstream Bandwidth Allocation Map (referred to only as MAP)
Mbps
Megabits per second
MCM
Multi-channel MPEG
M-CMTS
Modular Cable Modem Termination System
MER
Modulation Error Ratio
MHA
Modular Headend Architecture
MHz
Megahertz
MIB
Management Information Base
MKA
MACsec Key Agreement (protocol)
M/N
Relationship of integer numbers M,N that represents the ratio of the downstream symbol
clock rate to the DOCSIS master clock rate
MPEG
Moving Picture Experts Group
MPEG-TS
Moving Picture Experts Group Transport Stream
MPT
MPEG-TS mode of R-DEPI
MPTS
Multi Program Transport Stream
ms
Millisecond
MSB
Most Significant Bit
MSK
Master Secret Key
MSO
Multiple System Operator
MSB
Most Significant Bit
MTU
Maximum Transmission Unit
NAD
Network Access Device
NDF
Narrowband Digital Forward
NDR
Narrowband Digital Return
ns
Nanosecond
NSI
Network Side Interface
ONU
Optical Network Unit
OCSP
Online Certificate Status Protocol
06/15/15
CableLabs®
19
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
OSSI
Operations System Support Interface
PAE
Port Access Entity
PAT
Program Association Table
PCR
Program Clock Reference
PHB
Per Hop Behavior
PHB-ID
Per Hop Behavior Identifier
PHS
Payload Header Suppression
PHY
Physical Layer
PID
Packet Identifier
PKI
Public Key Infrastructure
PMD
Physical Media Dependent Sublayer
PMT
Program Map Table
PMTUD
Path MTU Discovery
PNM
Proactive Network Maintenance
PPP
Point-to-Point Protocol
PSI
Program Specific Information
PSIP
Program and System Information Protocol
PSP
Packet Streaming Protocol
PTP
Precision Time Protocol
PW
Pseudowire
QAM
Quadrature Amplitude Modulation
RDC
Regional Data Center
R-DEPI
Remote Downstream External PHY Interface
RF
Radio Frequency
RFI
Radio Frequency Interface
RFC
Request For Comments
RPD
Remote-PHY Device
RSA
Rivest-Shamir-Adleman (cryptosystem)
R-UEPI
Remote Upstream External PHY Interface
SCCRN
L2TPv3 Start-Control-Connection-Connected message
SCCRP
L2TPv3 Start-Control-Connection-Reply message
SCCRQ
L2TPv3 Start-Control-Connection-Request message
S-CDMA
Synchronous Code Division Multiple Access
SGID
Service Group Identifier
SLI
L2TPv3 Set Link Info message
SPTS
Single Program Transport Stream
SSD
Secure Software Download
StopCCN
L2TPv3 Stop-Control-Connection-Notification message
TCP
Transmission control protocol
TFTP
Trivial File Transfer Protocol
TSID
MPEG2 Transport Stream Identifier
20
CableLabs®
06/15/15
Remote PHY Specification
UCD
Upstream Channel Descriptor
UDP
User Datagram Protocol
US
Upstream
UTC
Coordinated Universal Time
VoD
Video On Demand
VoIP
Voice over IP
06/15/15
CM-SP-R-PHY-I01-150615
CableLabs®
21
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
5 TECHNICAL OVERVIEW
5.1
Introduction
In a Remote PHY Architecture, the classic integrated CCAP (I-CCAP) is separated into two distinct components. The
first component is the CCAP-Core and the second component is the Remote PHY Device (RPD).
The CCAP-Core contains both a CMTS Core for DOCSIS and an EQAM Core for Video. The CMTS Core contains
the DOCSIS MAC and the upper layer DOCSIS protocols. This includes all signaling functions, downstream and
upstream bandwidth scheduling, and DOCSIS framing. The DOCSIS functionality of the CMTS Core is defined by
[MULPI v3.0]. The EQAM Core contains all the video processing functions that an EQAM provides today.
The Remote PHY Device is a physical layer converter whose functions are:
•
To convert downstream DOCSIS, MPEG video and OOB signals received from a CCAP-Core over a digital
medium such as Ethernet or PON to analog for transmission over RF or linear optics.
•
To convert upstream DOCSIS, and OOB signals received from an analog medium such as RF or linear optics to
digital for transmission over Ethernet or PON to a CCAP-Core.
The RPD platform contains mainly PHY related circuitry, such as downstream QAM modulators, upstream QAM
demodulators, together with pseudowire logic to connect to the CCAP-Core.
It provides a subset of the following external interfaces:
CIN Facing:
•
One or more 10G or 1G Ethernet or PON ports
Access Network Facing:
•
One or more 10G or 1G Ethernet or PON ports
•
•
Additional RPDs may be daisy-chained through these ports
One or more RF ports providing connectivity to the access network
•
•
RF ports may be unidirectional (for use with an external combiner) or bi-directional (internal combiner)
RF port output may be RF over coaxial cable or over analog optics
An example reference implementation based on Ethernet is shown in Figure 1.
Figure 1 - Logical View of RPD Internals
The DOCSIS functionality of the Remote PHY Device is defined by[PHY v3.1], [MULPI v3.1], [DRFI], and [PHY
v3.0].
22
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
Together, the CCAP-Core and the RPD are the functional equivalent of an I-CCAP (Integrated CCAP), just with
different packaging. The MHAv2 specifications describe how the CCAP-Core and the RPD interface with each other.
Note that MHAv2 functionality and signaling and DOCSIS functionality and signaling are completely separate. The
DOCSIS functionality and signaling remain the same for both I-CMTS and Remote PHY solutions. MHAv2 focuses
on a simple deconstruction of the CMTS that moves the CCAP PHY elements into an external RPD device while
keeping the DOCSIS CMTS-to-CM signaling untouched.
5.2
System Diagram
Figure 2 shows an abstracted view of a cable operator’s network. Note that there are more aggregation points beyond
a headend such as a super headend or a regional data center (RDC). For the scope of this specification, the focus will
be on the headend aggregation point.
Figure 2 - Remote PHY System Diagram
The items in green are physical locations. The headend is where the majority of the equipment that does not require
direct connectivity to the access network resides. Video channel line-ups are often created in the headend. The
headend aggregates a number of hubs, with the hub containing equipment that requires direct connectivity to the HFC
plant. One example of equipment in the hub is the I-CCAP. The hub aggregates a number of optical nodes. The optical
nodes are located in the field and convert between a long point-to-point optical run and a local coax network. The
optical node aggregates traffic from a number of subscriber endpoints such as DOCSIS CMs and Video STBs.
5.2.1
Hub Access Network
The hub access network is the network that connects the headend and the hub. The hub access network can be either a
switched Layer 2 network or a routed Layer 3 network. It typically is a multi-hop network, which means there can be
multiple switches and/or routers between equipment in the headend and the hub.
5.2.2
Optical Access Network
The optical access network is located between the hub and the optical node. The access network has a forward path and
a reverse path.
5.2.2.1
Using Linear Optics
The classic HFC plant uses linear optics where the RF spectrum from the coax is modulated onto an optical
wavelength. The only type of signal that can traverse this type of network is an RF modulated signal such as a QAM or
an OFDM signal.
06/15/15
CableLabs®
23
CM-SP-R-PHY-I01-150615
5.2.2.2
Data-Over-Cable Service Interface Specifications
Using Digital Optics Only
A variation of the classic HFC plant uses digital optics in the return path. The RF spectrum is digitized and sampled at
the optical node, sent to the headend, and then reconstructed into an analog signal. From the viewpoint of this
specification, this will be considered as a subset of a linear optics HFC plant since its operation is transparent to the
transmission path which is still a modulated signal such as QAM or OFDMA.
5.2.2.3
Using Digital Optics with IP
A new HFC plant architecture is available that can use any fiber compatible baseband networking technology, such as
Ethernet, EPON, or GPON, to drive the fiber portion of the HFC plant. The coax portion of the HFC plant remains the
same. With digital optics based upon IP networking, the optical access network could be directly connected from the
CCAP-Core to the optical node. Since the hub is aggregating many optical nodes, the access network may have one or
more network elements in it, where the network elements could be a Layer 2 switch or a Layer 3 router. Note that this
model includes network elements that may be physically located at the hub but are connected between the CCAP-Core
and the optical node.
One of the goals of MHAv2 is to accommodate this new digital IP-based HFC plant architecture while maintaining the
minimum impact on the CCAP definition and operation. In this manner, I-CCAP and Remote PHY implementations
may be used as needed for different HFC plant architectures while maintaining a common CCAP feature set and
software loads.
5.2.3
Coax Access Network
The coax portion of the network is an FDM (frequency division multiplex) plant that carries RF modulated signals. It
has an upper frequency bound and a frequency range that is split between the upstream and downstream spectrums.
5.2.4
Location of the Remote PHY Device
For an optical access network based on linear optics, the RPD is located at the hub. For an optical access network
based on digital optics, the RPD is located at the optical node.
5.2.5
Location of the CCAP-Core
The previous version of I-CCAP is located at the hub where the RF ports can have direct connectivity to the access
network. Since the CCAP-Core does not have RF ports, this restriction is removed. The CCAP-Core can be located at
the hub or headend (or at another location beyond the headend, like the regional data center).
The network between the CCAP-Core and the RPD is known as the Converged Interconnect Network (CIN). The CIN
encompasses either or both the hub access network and the optical access network. The CIN can contain both Layer 2
switches and Layer 3 routers.
5.3
System Architecture
5.3.1
System Components
The reference architecture for a Modular CMTS system is shown in Figure 3. Architectures for video and OOB are
similar. This architecture contains both physical and logical components. This section briefly introduces each device
and interface.
24
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
Figure 3 - MHAv2 Reference Architecture for DOCSIS Signaling and Provisioning
The RPD is a component that has network interface on one side and an RF interface on the other side. The RPD
provides Layer 1 PHY conversion, Layer 2 MAC conversion, and Layer 3 pseudowire support. The RPD RF output
may be RF combined with other overlay services such as analog or digital video services.
The CCAP-Core contains everything a traditional CMTS does, except for functions performed in the RPD. The
CCAP-Core contains the downstream MAC, the upstream MAC, and all the initialization and operational
DOCSIS-related software.
Note that the original MHAv1 architecture had the downstream PHY external and the upstream PHY internal. MHAv1
was used to interface to an EQAM (Edge QAM) device that was co-located at the headend with the CMTS Core. Thus,
the main difference between MHAv1 and MHAv2 is the location of the upstream PHY and the role of the solution in
the marketplace. From a technical standpoint, the solutions are very similar.
Due to the physical separation of the downstream PHY and the upstream PHY in MHAv1, a DOCSIS Timing
Interface (DTI) Server was needed to provide a common frequency of 10.24 MHz and a DOCSIS timestamp between
the two MHAv1 elements. In MHAv2, the same DTI server is not required since the downstream and upstream PHYs
are co-located in the RPD. A different timing solution referred to as R-DTI is used to provide timing services for
functions such as DOCSIS scheduling.
R-DEPI, the Remote Downstream External PHY Interface, is the downstream interface between the CCAP-Core and
the RPD. More specifically, it is an IP pseudowire between the MAC and PHY in an MHAv2 system that contains
both a data path for DOCSIS frames, video packets, and OOB packets, as well as a control path for setting up,
maintaining, and tearing down sessions. MHAv1 used the MPT (MPEG-TS) encapsulation. MHAv2 retains the
original MPT encapsulation for backward compatibility but also added a new MPEG encapsulation called MCM
(Multi-channel MPEG). MHAv2 also requires the PSP (Packet Streaming Protocol) mode for expansion of new
services like DOCSIS 3.1.
R-UEPI, the Remote Upstream External PHY Interface, is the upstream interface between the RPD and the
CCAP-Core. Like R-DEPI, it is an IP pseudowire between the PHY and MAC in an MHAv2 system that contains both
a data path for DOCSIS frames, and a control path for setting up, maintaining, and tearing down sessions.
NSI, or the Network Side Interface, is unchanged, and is the physical interface the CMTS uses to connect to the
backbone network. Today, this is typically 10 Gbps Ethernet.
CMCI, or Cable Modem to Customer Premise Equipment Interface, is also unchanged, and is typically Ethernet,
USB, or WiFi. Within this document, CMCI is referred to as RPD.
06/15/15
CableLabs®
25
CM-SP-R-PHY-I01-150615
5.4
Data-Over-Cable Service Interface Specifications
Remote PHY Device Architecture
Figure 4 shows the architecture for an RPD.
Figure 4 - Remote PHY Device Block Diagram
Packet traffic arrives from the CCAP-Core on the downstream receiver. The DEPI framing is terminated; the payload
is extracted, framed, modulated, and transmitted out the cable interface. In the upstream, the signal is received from
the coax, digitized, demodulated, and the DOCSIS frames are extracted from the FEC payload. The DOCSIS frames
are then placed into the UEPI encapsulation and transmitted out the upstream transmitter to the CCAP-Core. A
clocking circuit interfaces to R-DTI and manages clocking and timing accuracy for the RPD. There is a local CPU that
manages the DEPI and GCP control planes and provides an interface into network management.
Figure 4 is meant to be explanatory and is not meant to be all-inclusive. Specific implementations may differ.
5.5
Remote PHY Operation
Figure 5 shows the internal components of an RPD. The following subsections explain the behavior and functionality
of these internal components.
Figure 5 - R-PHY Internal Components
5.5.1
R-DEPI and R-UEPI
R-DEPI and R-UEPI are IP-based pseudowires that are inserted between the DOCSIS MAC in the CCAP-Core and
the DOCSIS PHY in the RPD. R-UEPI is an extension to R-DEPI. R-UEPI uses the same control plane structure and
a unique set of encapsulations in the upstream direction.
R-DEPI’s job is to take either of the formatted DOCSIS frames, transport them through a Layer 2 or Layer 3 network,
and deliver them to the RPD for transmission. R-UEPI’s job is to take DOCSIS frames that have been received and
26
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
demodulated by the DOCSIS upstream PHY in the RPD and transport them to the CCAP-Core for processing. The
RPD does not provide any upstream DOCSIS processing; with one minor exception, the RPD will extract the
bandwidth request frames from the DOCSIS stream and send them in a separate pseudowire so that bandwidth request
frames can be given a higher priority than data frames.
The base protocol that is used for the R-DEPI is the Layer 2 Tunneling Protocol version 3, or L2TPv3 for short, and is
specified in document [RFC 3931]. L2TPv3 is an IETF protocol that is a generic protocol for creating a pseudowire,
which is a mechanism to transparently transport a Layer 2 protocol over a Layer 3 network. Examples of protocols
supported by L2TPv3 include ATM, HDLC, Ethernet, Frame Relay, PPP, etc.
Each data packet contains a 32-bit session ID. In the original MPT encapsulation, that session ID is associated with a
single QAM Channel. The UDP header—as part of an L2TPv3 encapsulation—is not used in the MHA protocols. The
L2TPv3 session ID directly follows the IP header. It is worth noting that the L2TPv3 session ID lands in the same part
of a packet as a classic UDP DP/SP. This allows network equipment that classify based upon UDP headers, to be
reused for L2TPv3 headers.
L2TPv3 permits creating a subheader whose definition is specific to the payload being carried. The control channel
allows for signaling messages to be sent between the CCAP-Core and the RPD. Typical control messages will set up a
"control connection" between the CCAP-Core and the RPD, and then set up multiple data sessions (one for each
downstream and upstream QAM or OFDM channel). Each session can be marked with different Differentiated
Services Code Points (DSCPs) and can support different encapsulation protocols.
There are two main pseudowire techniques defined by R-DEPI. Each main type supports a variety of subtypes. The
first technique, known as MPT mode, transports multiple 188-byte MPEG-TS packets by placing them into the
L2TPv3 payload with a unique subheader that contains a sequence number so packet drops can be detected. The
encapsulation of DOCSIS frames into MPEG-TS packets is performed in the CCAP-Core. The second technique,
known as the Packet Streaming Protocol (PSP), transports DOCSIS frames in the L2TPv3 payload. The DOCSIS
frames are then encapsulated in MPEG-TS packets within the EQAM. PSP mode allows DOCSIS frames to be both
concatenated, to increase network performance, and fragmented, in case the tunneled packets exceed the network
MTU size. MPT mode is generally used for single carrier QAM systems such as DOCSIS 3.0 and video, while PSP
mode is used for downstream OFDM channels and for the DOCSIS upstream.
5.5.2
Remote DTI
Remote DTI (see [R-DTI]) provides timing synchronization between CCAP-Cores and RPDs based on the IEEE
1588v2 standard [IEEE 1588]. The protocol supports the basic synchronization between the CCAP-Core and Remote
PHY Device for DOCSIS/video/OOB services and the precision time synchronization for emerging services such as
wireless backhaul.
5.6
Latency
One of the technical considerations of the MHAv2 architecture is its impact on the round-trip request-grant delay time.
The request-grant delay time is the time from when a CM requests bandwidth, using an uncontended bandwidth
request (REQ), to when it receives a MAP message with the granted transmit opportunity in it.
MHAv2 locates the upstream scheduler in the CMTS Core. To prevent the MAP from being slowed down by other
traffic in the CIN, the DOCSIS traffic (or a subset containing the MAP messages) may be sent in an independent
L2TPv3 flow that can have a unique DSCP. The value of the marked DSCP value should be consistent with a
configured "per hop behavior (PHB)" that will provide MAP messages with the highest priority and lowest latency
across the CIN to the EQAM. Marking of the DHCP field is optional and part of the operator’s overall network design.
In the upstream direction, the request can be copied from the DOCSIS frame and sent on an independent L2TPv3 flow
that has a unique DSCP.
The net result of prioritizing the MAP and REQ messages, combined with a good CIN design, is to make the operation
and performance of the centralized upstream scheduler similar to that of an I-CMTS system.
06/15/15
CableLabs®
27
CM-SP-R-PHY-I01-150615
5.7
Data-Over-Cable Service Interface Specifications
MHAv2 Summary
In summary, the RPD is used to transfer DOCSIS frames between an IP network interface and an RF interface. The
RPD does not participate in the DOCSIS MAC protocol. Instead, MHAv2 provides an IP pseudowire that seamlessly
transports the DOCSIS frames between the CCAP-Core and the RPD. As such, for most DOCSIS functions, the
MHAv2 CCAP system functions almost identically to an I-CCAP. This preserves common functionality and features
between the two systems.
28
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
6 RPD INITIALIZATION
6.1
Overview
When the RPD device first powers up, it goes through a series of steps before becoming operational. These steps are
shown in Figure 6 and explained in this section. Note that Figure 6 is a simplified sequence and does not attempt to
show error paths. Error handling during each operation is described in a subsequent, relevant section of the document.
Figure 6 - RPD Initialization
06/15/15
CableLabs®
29
CM-SP-R-PHY-I01-150615
6.2
Data-Over-Cable Service Interface Specifications
Security
The Remote PHY security architecture consists of a trusted domain and an untrusted domain (see Figure 7 below). To
access the trusted domain and connect to the CCAP-Core, RPDs may be required to be authenticated to the trusted
network. This is accomplished using 802.1x. When the RPD connects to the CCAP-Core, a control session is
established which can be secured using IPsec. Both of these mechanisms perform mutual authentication using digital
certificate credentials issued from a trusted public key infrastructure (PKI). RPDs support both of these mechanisms
(802.1x or IPsec). MSOs can enable them as needed for their specific deployments.
Details for both mechanisms are provided in the following sections.
6.3
6.3.1
Network Authentication
Problem Definition
In many cases, an RPD will be located in an untrusted part of the MSO network, such as a pole-mounted fiber node or
remote cabinet but must connect to devices inside the trusted network. When this occurs, it presents a potential
security vulnerability. An RPD in an environment like this is shown in Figure 7.
Figure 7 - Remote PHY: Trusted Domain and Untrusted Domain
To mitigate this threat, an MSO may require that the RPD is authenticated before it is allowed access to the trusted
network. An RPD may of course be located within the trusted network boundary, such as in a physically secured hub
site. In this case, authentication may not be required. Thus, the RPD must be able to operate in both authenticated and
unauthenticated networks. Whether authentication is required for an RPD is determined by the network that it is
connected to rather than the RPD itself. To support “out of box” operation, an RPD should first attempt to authenticate
to the network. If no response to authentication is received, it should assume authentication is not supported by the
network and attempt to operate without it (refer to Section 6.3.3.4 for details).
6.3.2
Authentication from an Untrusted Portion of the Network
In Figure 8, the RPD is located in an untrusted area of the network so the network is configured to require
authenticated access. The CCAP-Core is located in a trusted area of the network. A single RPD may connect to more
30
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
than one CCAP-Core. This is because there may be different CCAP-Cores for DOCSIS and video, or for primary and
standby. The RPD will also need to connect to other network services such as DHCP and to allow connections from
network management servers.
Figure 8 - Authentication Network Diagram
The two authentication scenarios that an RPD MUST support are:
•
No authentication in which case the RPD can send to and receive packets from the trusted network with no
additional requirements.
•
802.1x based authentication, which requires the RPD to act as an 802.1x supplicant, as described in Section 6.3.3.
A fundamental objective for deployment of RPD is to not require configuration.
To achieve this “out of the box” operation, the RPD MUST be able to determine which security option is in place
without configuration.
The RPD MUST determine whether 802.1x authentication is operating, as described in Section 6.3.3.
6.3.3
802.1x Authentication
Authentication is performed based on the 802.1x [IEEE 802.1x] and MACsec [IEEE 802.1ae] standards.
802.1x is a Layer 2 protocol that uses EAP (Extensible Authentication Protocol) to provide authentication services.
For the RPD, EAP-TLS is used based on digital certificate credentials issued from the DOCSIS PKI.
The standard defines three entities:
Supplicant
This is the RPD that requires authentication.
Authenticator/NAD
This is a network element that prevents the RPD from gaining network
access until authentication is achieved. The Authenticator is also known as
a Network Access Device (NAD).
Authentication Server
This is a standard 802.1x authentication server that validates the
authentication.
MHAv2 uses a standard version of the 802.1x protocol with EAP-TLS. This method is referred to as network
authentication since the entire authentication process happens between the RPD and a Network Access Device (NAD)
without the involvement of the CCAP-Core.
06/15/15
CableLabs®
31
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
Figure 9 shows how the EAP messages between the Authentication Server and the Authenticator are carried over the
Radius or Diameter, while the EAP messages from the Authentication Server to the RPD are carried over the
combination of Radius/Diameter and 802.1x (EAPoL).
Figure 9 - Network Authentication Signaling
The Authenticator will transmit a Layer 2 broadcast EAP-Request message periodically or in response to an
EAPoLstart message from a supplicant. An RPD will respond with a Layer 2 unicast EAP-Response. The
Authenticator will forward the EAP response to the Authentication server using a RADIUS or DIAMETER protocol.
The Authentication server and the RPD then communicate directly using the Authenticator as a relay agent. When the
Authentication server has made a decision, it communicates that decision to the Authenticator. The Authenticator will
then provide or deny network access to the RPD.
6.3.3.1
MACsec
MACsec (see [IEEE 802.1ae] is a link layer encryption mechanism used to provide additional security to 802.1x.
MACsec may be used to provide link level encryption between the RPD and the NAD. A security association is
created between the NAD and each authenticating RPD based on keying material created during the EAP exchanges.
This is used to encrypt data between the NAD and each RPD, providing a higher level of security than basic 802.1x.
With 802.1x, after authentication of an RPD, the NAD port is opened to any messages from the authenticated RPD
MAC address. This creates the possibility for a device to spoof the RPD MAC address to gain access to the network.
With MACsec only devices in possession of legitimate security keys can send traffic to the network.
The use of MACsec provides the following advantages:
•
It enables secure access for multiple devices per port
•
It provides protection against potential man in middle attacks in both single and multiple devices per port use
cases.
MACsec is relatively new and is not yet supported by all switches and all silicon. If MACsec is supported
•
The RPD MUST support the MACsec Key Agreement protocol (MKA) for key exchange and management.
•
The RPD MUST derive the Connectivity Association Key (CAK) from the EAP-MSK as defined by EAP-TLS
and 802.1x.
•
The RPD MUST not use pre-shared CAKs.
32
CableLabs®
06/15/15
Remote PHY Specification
6.3.3.2
CM-SP-R-PHY-I01-150615
RPD Topology Support for 802.1x
Figure 10 shows a number of potential topologies for RPD deployment and connectivity to the NAD. The various
topologies are discussed in the subsections of this section.
Figure 10 - RPD Topologies for 802.1x
6.3.3.2.1
Type 1: Single Host per Port
In its most basic form, 802.1x supports access by a single host per Ethernet switch port. This is defined in the 802.1x
standard and is widely supported by existing switches.
The RPD MUST support this topology.
The RPD MUST support 802.1x in this configuration.
The RPD SHOULD support MACsec in this configuration.
6.3.3.2.2
Type 2: Daisy Chained RPDs
In this topology, a single NAD port is connected to multiple RPDs connecting over a single port. In this topology
MACsec may be used to establish independent security associations with each RPD. This is defined in the standard but
is not widely supported in current switches.
The RPD MAY support this daisy chain topology.
If a daisy chain topology is supported:
•
The RPD SHOULD support 802.1x in this configuration;
06/15/15
CableLabs®
33
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
•
The RPD SHOULD support MACsec in this configuration.
•
802.1x request messages are carried in a multicast packet with the well-known PAE group address as the
destination address. Normal Ethernet switches are required to block this address so that 802.1x is typically a
single hop protocol with the authenticator directly connected to the supplicant.
•
The RPD SHOULD propagate the 802.1x EAP-REQ multicast messages between the NAD and the daisy chain
port.
6.3.3.2.3
Type 3: Multiple RPDs in Single Device
In this topology, a single NAD port is connected to an integrated device such as a node with multiple RPDs connecting
via an internal hub or switch. Thus the NAD sees multiple devices (and multiple MAC source addresses) on the port.
In this topology, MACsec may be used to establish independent security associations with each RPD, based on the
RPD MAC address. This is defined in [IEEE 802.1ae], but is not widely supported in current switches.
If this topology is supported:
•
Each RPD MUST have a unique MAC address per Ethernet port;
•
The internal hub / switch SHOULD propagate the 802.1x EAP-REQ multicast messages to the RPDs;
•
The RPD SHOULD support 802.1x in this configuration;
•
The RPD SHOULD support MACsec in this configuration.
6.3.3.2.4
Type 4: Intermediate External Switch / Router
In this topology, a one or more RPDs are connected to the NAD through an intermediate switch or router.
MSOs deploying this topology may not be able to utilize 802.1x or MACsec due to the forwarding restrictions on PAE
multicast in Ethernet bridges and switches. Definition of the external switch behavior that would be required to
support this topology is outside the scope of this specification.
6.3.3.3
Authenticator Location
The Authenticator is hosted in the device at the border of the trusted network. This may be a Layer 2 switch, a Layer 3
router or the CCAP-Core.
There can be zero or more Layer 2 switches and/or Layer 3 routers between the Authenticator and the CCAP-Core.
There can be zero or more Layer 2 switches and/or Layer 3 routers between the Authenticator and the Authentication
Server.
6.3.3.4
Operation
After powering up (and prior to obtaining an IP address) the RPD MUST attempt to authenticate itself to the network
using 802.1x as shown in Figure 11.
The RPD MUST send an EAPOL-START message to the Authenticator and wait for an EAP-REQ. If there is no
EAP-REQ in response to the EOPOL-Start within “EAP-REQ-TIMEOUT”, the EAPOL-START MUST be resent and
the RPD MUST return to wait mode. If no EAP-REQ is received after EAPOL-START-RETRIES have been
exhausted, the RPD MUST assume that the network is not authenticated, operate in a non-authenticated mode, and
proceed with the DHCP phase of the initialization sequence (this is standard operating procedure for an 802.1x
device).
If an EAP-REQ is received the RPD MUST proceed with 802.1x authentication. If the RPD authentication is rejected
or if the authentication process fails after an EAP-REQ has been received, the RPD MUST hold off for the defined
802.1x wait period before trying to re-authenticate.
Once authentication is completed succesfully, the RPD MUST proceed with the DHCP phase of the initialization
sequence.
34
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
Figure 11 - RPD Authentication using 802.1x
6.3.3.5
802.1x Mutual Authentication
802.1x with EAP-TLS provides mutual authenticaton of the RPD and the Authentication server. The RPD MUST use
EAP-TLS per [RFC 5216] with certificates issued from the DOCSIS PKI managed by CableLabs (see Annex E). The
CableLabs Root CA certificate is installed in the Authentication server and RPD as a trust anchor for validating
received certificates. The RPD Certificate and its private key, along with the issuing intermediate Device CA
certificate are installed in the RPD. The Authentication Server Certificate and its private key, along with the issuing
06/15/15
CableLabs®
35
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
intermediate Service Provider CA certificate, are installed on the Authentication server. During the EAP-TLS message
exchange, the RPD and Authentication server will send their device/server certifcates and the issuing intermediate CA
certificate to each other to be validated against the root CA trust anchor certificate. The RPD and Authentication server
MUST use the “Basic Path Validation” procedure defined in [RFC 5280] for validating received certificates.
6.3.3.6
CCAP-Core Requirements
The CMTS Core MAY act as a NAD if it is directly connected to the RPD. In ths case, it MUST support the 802.1x
protocol and act as a relay agent to the Authentication server.
6.3.3.7
Authentication Failures
If an EAP-REQ message has been received indicating that authentication is in effect for the network, any subsequent
failures during the authentication process MUST be handled per the [IEEE 802.1x] specification. The wait period
timer (which defines the time a device must wait after a failed authentication attempt before another attempt is
permitted) SHOULD not be reduced below the 60 second default time.
Retransmission behavior for EAP messages (which are forwarded from the authenticator to the authentication server)
MUST follow [RFC 3748].
6.4
Constant
Value
EAP-REQ-TIMEOUT
10 sec
EAPOL-START-RETRIES
3
Address Assignment
Figure 12 shows a simplified access network containing an RPD, a CCAP-Core and a DHCP Server.
Figure 12 - DHCP Network Diagram
Once the RPD is successfully authenticated to the trusted network, it requests an IP address using DHCP. The standard
DHCPv6 or DHCPv4 protocol is used with extensions.
The RPD MUST initially run DHCPv6. If that fails, then the RPD MUST run DHCPv4.
36
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
This method is referred to as network DHCP since the entire address assignment of the RPD can take place without the
involvement of the CCAP-Core. The CCAP-Core MAY run a DHCP Relay agent but is not required to if relay is
provided by another network component.
There may be zero or more Layer 2 switches between the RPD and the DHCP Relay. The DHCP Relay may be hosted
by a Layer 2 switch, a Layer 3 router or the CCAP-Core. There may be zero or more Layer 2 switches and/or Layer 3
routers between the network element that is hosting the DHCP Relay and the CCAP-Core. There may be zero or more
Layer 2 switches and/or Layer 3 routers between the DHCP Relay and the DHCP Server.
Figure 13 shows the DHCP signaling protocol. The RPD issues a broadcast DHCP discovery message when it needs to
obtain an IP address. The DHCP agent responds with a unicast DHCP offer that contains the IP address of one or more
DHCP servers. The RPD picks one of the DHCP servers and sends a DHCP request to it. The DHCP server sends a
DHCP acknowledgement with an IP address for the RPD device.
Figure 13 - DHCP Signaling
Unlike the DOCSIS DHCP process where the CCAP is always the DHCP relay agent and can always snoop and
append to the DHCP messages, the CCAP-Core may not have any direct access to the DHCP message exchange and
thus will not be directly aware of the IP address assignment of the RPD. The following mechanism MUST be used to
create an association between the CCAP-Core and the RPD:
1.
The CCAP-Core and the DHCP Server MUST be provisioned with the same IP/MAC Address pair for the RPD.
The provisioning can be done manually or with an automated system.
2.
The DHCP Server MUST provide the IP address of the CCAP-Core to the RPD. An additional DHCP option
<CCAP-Cores> is added to support this mechanism. The CCAP-Core MUST either accept the connection from
the RPD, deny the connection, or redirect the RPD to another CCAP-Core.
6.4.1
DHCP Options
Refer to [CANN] for details on specific options.
The RPD MUST support the following DHCP options when they are received in a DHCP message.
Option
Value
Use
2
Time Offset
Used for authentication, logging, and software upgrade
4
Time Server
Used for authentication, logging, and software upgrade
7
Log Server
Used for logging
The RPD MUST support the following CableLabs suboptions under DHCP option 43.
Suboption
2
06/15/15
Value
<Device Type>
Use
MUST be set to “RPD”
CableLabs®
37
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
3
<ECM: eSAFE>
Not used
4
<serial number>
Refer to [CANN]
5
<hw version>
Refer to [CANN]
6
<sw version>
Refer to [CANN]
7
<Boot ROM version>
Refer to [CANN]
8
<OUI>
Refer to [CANN]
Refer to [CANN]
9
<Model Number>
10
<Vendor Name>
Refer to [CANN]
TBD
<CCAP-Cores>
Address of all CCAP-Cores RPD MUST attempt to connect to. The Principal Core is the
first entry in the list.
6.4.1.1
CCAP-Cores Suboption
CMTS Cores suboption describes either IPv4 or IPv6 addresses, as shown in Figure 14 and Figure 15.
Figure 14 - CCAP-Cores DHCP Suboption IPv4
The CMTS Cores suboption can also be used with DHCPv6.
Figure 15 - CCAP-Cores DHCP Suboption IPv6
The cores may have different roles, such as primary, standby, DOCSIS, EQAM, etc. The specific role of each core is
determined during the GCP configuration phase.
•
The RPD MUST attempt to connect to all cores in the options list as described in Section 6.6.
6.4.2
Failures
The RPD MUST respond to any errors during the DHCP process as per [RFC 2131].
NOTE: This results in the RPD entering a time out and retry loop with a randomized exponential back off.
38
CableLabs®
06/15/15
Remote PHY Specification
6.4.3
CM-SP-R-PHY-I01-150615
Security Implications
The RPD MUST attempt to contact the DHCP server via the CIN interface. If 802.1x and MACsec are in place this
will provide secure access to the trusted network.
6.5
Time of Day
The RPD acquires the time of day for the purpose of timestamping warnings, error logs and messages, validation of the
CVC during a software upgrade, and validation of the CCAP-Core certificate during mutual authentication.
6.5.1
ToD Acquisition
The RPD MUST attempt to obtain the current date and time by using the Time Protocol (see [RFC 868]) from one of
the servers listed in the Time Server Option DHCP field. If this field is missing or invalid, the RPD MUST initialize
the current time to Jan 1, 1970, 0h00. If the time is initialized (reset), the RPD MUST ignore the value, if any, of the
Time Offset DHCP option.
The RPD MUST use its DHCP-provided IP address for exchange of messages with the Time Protocol server. The
RPD MUST transmit the request using UDP. The RPD MUST listen for the response on the same UDP port as is used
to transmit the request. The RPD MUST combine the time retrieved from the server (which is UTC) with the time
offset received from the DHCP server to create a valid "local" time.
Once the RPD acquires time, it MUST stop requesting, unless any of its ToD related parameters (such as time offset or
server address) are modified. If the RPD’s ToD related parameters are modified, the RPD MAY re-request ToD from
the Time Protocol server(s).
6.5.2
ToD Conflicts and Problems
The DHCP server may return multiple IP addresses from multiple Time Protocol servers. The RPD MUST attempt to
obtain time of day from all the servers listed until it receives a valid response from any of the servers. The RPD MUST
contact the servers in batches of tries with each batch consisting of one try per server and each successive try within a
batch at most one second later than the previous try and in the order listed by the DHCP message. If the RPD fails to
acquire time after any batch of tries, it MUST retry a similar batch using a truncated randomized binary exponential
backoff with an initial backoff of 1 second and a maximum backoff of 256 seconds.
If an RPD is unable to establish time of day it MUST log the failure in the local log. If the RPD does not obtain ToD in
the initial request against the first server, the RPD MUST initialize the current time to Jan 1, 1970, 0h00, and then
subsequently initialize its current time once it receives a response from a Time Server.
If the RPD fails to establish TOD it will not be able to validate the CCAP-Core certificate.
The failure may be due to time server failure or to an error in the DHCP option list.
The RPD will follow the same mechanism as for a DHCP failure and restart the DHCP process as described in Section
6.4.2
6.5.3
ToD Security Implications
The RPD MUST attempt to contact the time server via the CIN-facing port on which the DHCP response was
received. If 802.1x and MACsec are in place this will provide secure access to the trusted network.
6.6
Connection to CCAP-Cores
Following successful IP address assignment the RPD attempts to connect to all of the CCAP-Cores that have been
identified in the DHCP option list as described below.
6.6.1
Core Types
Cores are defined to be either Principal or Auxiliary.
06/15/15
CableLabs®
39
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
An RPD can be connected to multiple CCAP-Cores. Each CCAP-Core manages and configures an independent subset
of the RPD resources, e.g., one or more RF channels. There are certain types of parameters, which are common across
resource sets such as downstream power. The Principal Core is responsible for the configuration of these common
parameters for the RPD and for certain device management functions.
Auxiliary cores are responsible for providing DOCSIS, video, or OOB services. They are restricted to the resource set
assigned to them by the Principal Core.
The RPD MUST complete configuration with a Principal Core before allowing configuration from Auxiliary cores.
In general, it is expected that the first core in the DHCP option list will be the Principal, but the RPD MUST be able to
accommodate out-of-order lists.
The RPD MUST accept configuration from only one Principal core (refer to Section 6.6.2.4 for details).
Principal and Auxiliary cores may operate in Active or Standby roles. {Note 1}
Note 1: High Availability actions will be defined in a future version of the specification
6.6.2
Connection Process
The connection process between RPD and each CCAP-Core (whether Principal or Auxiliarry) consists of two phases:
1.
Establish a mutually authenticated secure connection between the RPD and Core.
2.
RPD configuration using GCP.
6.6.2.1
Mutual Authentication
Mutual authentication is used when establishing a secure connection between the RPD and CCAP-Core(s). It is
independent from the authentication used for trusted network access described in Section 6.3.
Mutual authentication is always required between the RPD and CCAP-Core but a secure connection may not be
required in all cases (e.g., when the RPD is inside the trusted network or MACsec is used to secure access to the trusted
network). This is negotiated as described below.
Authentication can be initiated by either the CCAP-Core or the RPD.
Whether the RPD is required to authenticate is under control of the CCAP-Core.
The RPD and the CCAP-Core MUST support mutual authentication based on IKEv2 [RFC 7296] using public key
signatures based on the digital certificate credentials issued from the CableLabs DOCSIS PKI (see Annex E). The
RPD certificate provisioning requirements are the same as what is defined in Section 6.3.3.5. The CCAP-Core MUST
be provisioned with the CableLabs Root CA certificate as a trust anchor. CCAP-Core MUST be provisioned with a
CCAP-Core Device Certificate and its private key along with the CableLabs Device CA certificate which are issued
by the CableLabs Root CA.
The RPD MUST use UDP port 500 for IKEv2 exchanges.
The CCAP-Core MUST use UDP port 500 for IKEv2 exchanges.
The RPD MUST initiate the IKE_SA_INIT process per [RFC 7296].
The CCAP-Core MAY initiate the IKE_SA_INIT process per [RFC 7296] if the RPD address is known in order to
facilitate faster reconnection (for example, following a CCAP error).
The RPD MUST attempt to recontact a CCAP-Core following a connection loss.
The RPD MUST include its [X.509] certificate in the IKEv2 exchanges.
The CCAP-Core MUST include its [X.509] certificate in the IKEv2 exchanges.
The RPD MUST use the value of the common name field from the [X.509] certificate as the identifier in IKEv2
messages it generates.
The CCAP-Core MUST use the value of the common name field from the [X.509] certificate as the identifier in IKEv2
messages it generates.
40
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
The mechanism by which the CMTS Core determines whether to accept the connection from the RPD is a local matter
but could include:
•
Local configuration of RPD identifier, or
•
Forwarding to an authentication policy server.
The CCAP-Core MUST determine the security profile for the GCP control plane during the IKEv2 exchange.
The GCP control plane SHOULD be authenticated by the CCAP-Core. The CCAP-Core MAY encrypt the GCP
Control Plane.
If authentication or encryption are in operation, IPsec ESP in transport mode MUST be used to protect the GCP
control plane.
If authentication or encryption are in operation, IKEv2 MUST be used to generate the keying material required to
secure the GCP control plane.
The IKEv2 traffic selector MUST be the 5-tuple identifying the GCP connection.
The CCAP-Core MUST determine the security profile for the L2TPv3 control plane during the IKEv2 exchange.
The L2TPv3 control plane MAY be authenticated and MAY be encrypted by the CCAP-Core.
If authentication or encryption are in operation, IPsec ESP in transport mode MUST be used to protect the L2TPv3
control plane.
If authentication or encryption are in operation, IKEv2 MUST be used to generate keying material to secure the
L2TPv3 control plane.
The IKEv2 traffic selector MUST be the 5-tuple identifing the L2TPv3 connection.
Different security associations MUST be used for GCP and L2TPv3 control.
The following cryptographic methods as defined in [RFC 4307] MUST be supported:
•
Message integrity using HMAC-SHA1-96;
•
Data Encryption using AES 128 CBC;
•
Pseudo-random function for key generation HMAC-SHA1;
•
Certificate authentication using RSA Signature Algorithm [RSA 3] with SHA-256 hash (see [FIPS 180-4]) per
Annex E.
6.6.2.1.1
No Authentication Option
If the CCAP-Core does not wish to use encryption or authentication, it signals this by selecting null encryption and no
authentication options (see [RFC 4307]) during the IKEv2 exchange.
6.6.2.1.2
Certificate Validation
The RPD MUST use the “Basic Path Validation” procedure defined in [RFC 5280] for validating received certificates.
The CCAP-Core MUST use the “Basic Path Validation” procedure defined in [RFC 5280] for validating received
certificates.
6.6.2.1.3
Authentication Failure
Failures during the authentication process MUST be handled per [RFC 7296].
If the authentication is terminated due to a failure, the RPD will attempt to connect to the next core in the list.
6.6.2.2
RPD Configuration via GCP
Following authentication, the CCAP-Core MUST configure the RPD using the GCP (Generic Control Plane) protocol
(see [GCP]). Since the RPD is an extension of the CCAP-Core, the CCAP-Core contains all the necessary
configuration information.
06/15/15
CableLabs®
41
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
GCP allows control plane data structures from other protocols to be tunneled through its generic control plane. For
example, GCP can directly use DOCSIS TLVs for the configuration of the RPD PHY parameters.
Note that the [R-DEPI] and [R-UEPI] protocols also contain a certain amount of configuration information. The
MHAv2 paradigm is to keep the R-DEPI and R-UEPI configuration focused on session signaling and to use GCP for
RPD-specific configuration and operation.
The specific RPD configuration parameters used in GCP are listed in Annex B.
The GCP protocol is authenticated and secured using IPsec. Encryption and/or message authentication codes (HMAC)
can be applied to protect packets. IPsec keys are derived from the keying material created during the IKEv2
authentication process. IPsec session key exchange and renewal during the life of the GCP connection will be
supported using IKEv2.
6.6.2.3
GCP Connection Failures
If a CCAP-Core has not responded after CORE_CONNECT_TIMEOUT, then the RPD MUST retry the connection
CONFIG_RETRY_COUNT. If no response is received after retries are exhausted, the RPD MUST move on to the
next Core in the list.
Constant
6.6.2.4
Value
CORE_CONNECT_TIMEOUT
5 seconds
CONFIG_RETRY_COUNT
3
Connection to Principal Core
Following successful IP address assignment, the RPD MUST follow the process shown in Figure 16 to connect to a
Principal core.
42
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
Figure 16 - Process for Connecting to the Principal Core
06/15/15
CableLabs®
43
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
The RPD MUST establish a GCP connection with a Principal Core following the process shown in Figure 16 and
described below. It MUST start the process with the first core in the DHCP option list and move sequentially through
the list until a successful connection is achieved.
The RPD MUST attempt to authenticate with the core as described in Section 6.6.2.1.
Following authentication, the RPD MUST initiate a TCP connection to the GCP well-known port on the CCAP-Core.
When the connection is established, the RPD MUST issue a GCP Notify message to the CCAP-Core to initiate
configuration and determine if the core can act as a Principal.
If the core identifies as a Principal core in active mode (as opposed to standby) the RPD MUST proceed with GCP
configuration.
If the configuration received from the Principal core overwrites any of parameter values communicated via the DHCP
Options previously received, the RPD MUST use the parameter values received from the Principal Core.
Alternatively, a new list of Auxiliary cores could be provided.
When configuration by the Principal core is complete, the RPD MUST initiate PTP clock synchronization. The PTP
may take some time to reach a steady state, so the RPD should start the synchronization process as soon as possible.
The RPD MUST establish L2TPv3 connectivity with the Principal core as described in Section 6.8.
Following successful L2TPv3 establishment, the RPD is operational with the Principal Core and MUST move on to
any Auxiliary cores defined by DHCP Options or defined during configuration from the Principal core.
6.6.2.5
Failures
If the core contacted is not a Principal Core, the RPD MUST move to the next core in the options list and attempt to
contact this core.
If the end of the option list is reached with no Principal Core found, the RPD MUST wait
NO_PRINCIPAL_CORE_FOUND_TIMEOUT then retry from the start of the list.
If no Principal Core can be contacted after PRINCIPAL_CORE_RETRY_COUNT attempts, the RPD must wait for a
random backoff time between PC_BACKOFF_MIN and PC_BACKOFF_MAX, and then reboot.
Constant
6.6.2.6
Value
NO PRINCIPAL_CORE_FOUND_TIMEOUT
60 seconds
PRINCIPAL_CORE_RETRY_COUNT
3
PC_BACKOFF_MIN
60 seconds
PC_BACKOFF_MAX
300 seconds
Redirection
If a Principal CCAP-Core does not have configuration data for an RPD or is not aware of the RPD, the core SHOULD
either reject the connection and log an error or use GCP to redirect the RPD to another core.
A CCAP-Core MAY elect to redirect an RPD to one or more alternate cores for further configuration, e.g., to act as a
standby or to provide additional services.
The CCAP-Core MUST use the GCP (Generic Control Plane) protocol to redirect the RPD.
The redirecting CCAP-Core MUST transfer a variable length list of IPv4 or IPv6 addresses to the RPD.
6.6.3
Connection to Auxiliary Cores
After becoming operational with a Principal core, the RPD MUST follow the process shown in Figure 17 to connect to
any Auxiliary cores that have been configured.
44
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
Figure 17 - Process for Connecting to Auxiliary Cores
For each core in the list:
•
If this is the active Principal core (to which it is already connected) the RPD MUST move to the next entry in the
list.
06/15/15
CableLabs®
45
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
•
The RPD MUST try to authenticate with the core.
•
Following authentication the RPD MUST initiate a TCP connection to the GCP well-known port on the
CCAP-Core. When the connection is established, the RPD MUST issue a GCP Notify message to the CCAP-Core
to initiate configuration.
•
If the core is an additional Principal core operating in active mode, the RPD MUST log an error and close the GCP
connection because only one active Principal core is allowed. The CCAP-Core MUST also log an error.
•
If the core is a Principal core operating in standby mode the RPD MUST retain this information in case the active
Principal core fails.
•
If the core is an auxiliary operating in standby mode the RPD MUST retain this information in case the active
Auxiliary core fails.
•
If the core is an auxiliary core operating in active mode:
•
•
•
6.7
The RPD MUST proceed with GCP configuration.
The RPD MUST establish L2TPv3 connectivity with the Auxiliary core as described in Section 6.8.
Following successful L2TPv3 establishment, the RPD is operational with the Auxiliary Core and MUST
move on to any additional Auxiliary cores defined by DHCP Options or defined during configuration
from the Principal core.
Synchronization
Once the RPD has been configured, the RPD chooses its method of synchronization. The RPD can be directed to either
be internally synchronized where the RPD is the clock master (Option A) or externally synchronized where the RPD is
a clock slave (Option B).
In a Remote PHY system, the downstream and upstream PHY timing are always aligned because the downstream and
upstream PHY are co-located. The only timing requirement is to be able to share a timestamp value between the
CCAP-Core and the RPD for upstream scheduling. These timing techniques are described in [R-DTI].
Note that if the upstream scheduler is located in the RPD, then all the timing elements are local to the RPD and no
adjustments are necessary. This scenario is equivalent (from a timing standpoint) to having the entire CMTS in the
RPD. This is a future option for the R-PHY architecture should it ever be needed.
The net effect of all methods is that the timestamp used in the SYNC message, the MAP message, the REQ message,
the RPD upstream burst receiver, and the upstream scheduler are aligned.
The protocol used in [R-DTI] between the CMTS Core and the RPD is the Precision Time Protocol (PTP) as defined
by [IEEE 1588]. PTP is used because it is a standard protocol whose accuracy can be enhanced when the CIN is built
with [IEEE 1588] compliant equipment. Note that it is not necessary for the network to be compliant to [IEEE 1588].
Encryption or authentication of PTP messages between the master clock and the RPD (e.g., by using IPsec) would
result in some loss of accuracy because intermediate nodes could not update the timing data. If security of PTP
messages is required, MACsec encryption can be used.
The synchronization requirements can be summarized as follows:
•
All specific operational requirements are stated in the [R-DTI] specification.
•
The RPD MUST be able to support PTP messages received over a MACsec (see [IEEE 802.1ae]) secured link
from the CIN.
•
If the CCAP-Core has specified a PTP Master source during GCP configuration the RPD MUST use it.
6.7.1
6.7.1.1
Synchronization Failures
RPD Operating as a Timing Slave
If the RPD does not receive a sync message within PTP_SYNC_TIMEOUT, it MUST log a LOSS_OF_SYNC error
and operate in a non-synchronized local clocking mode. It will continue to attempt to synchronize with a PTP clock
master. When synchronization is re-established, a SYNC_ESTABLISHED message MUST be logged.
46
CableLabs®
06/15/15
Remote PHY Specification
6.7.1.2
CM-SP-R-PHY-I01-150615
RPD Operating as a Timing Master
If the RPD does not receive a delay request message within PTP_DELAY_TIMEOUT, it MUST log a
LOSS_OF_SLAVE error. It will continue to act as PTP clock master. When communication with a slave is
re-established a SLAVE_FOUND message MUST be logged.
Constant
6.8
Value
PTP_SYNC_TIMEOUT
5 seconds
PTP_DELAY_TIMEOUT
5 seconds
Connectivity
Once clocking has been established, the RPD and the CCAP-Core are ready to set up the L2TPv3 data tunnel and
control plane connectivity.
Downstream data plane connectivity between the CMTS Core and the RPD is described in [R-DEPI]. Upstream data
plane connectivity between the RPD and the CMTS Core is described in [R-UEPI]. [R-DEPI] is an extension of the
Layer 2 Tunneling Protocol described in [RFC 3931].
R-DEPI and R-UEPI establish one overall control plane connection between a CMTS Core and an RPD pair. Within
this tunnel, there are separate pseudowires consisting of L2TPv3 sessions for each MAC-PHY functional pair. The
data plane encapsulation is managed per session. Depending upon the type of pseudowire encapsulation used, a
pseudowire may contain one or more channels.
After the R-DEPI and R-UEPI session initializes, the CMTS Core and RPD are ready to use.
The connectivity requirements can be summarized as follows:
•
If the RPD is provided with a Principal CCAP-Core IP address, the RPD MUST try to establish connectivity to
that CCAP-Core using the DEPI primary session.
•
If the RPD is provided with a Auxiliary CCAP-Core IP address, the RPD MUST try to establish connectivity to
that Auxiliary CCAP-Core using the auxiliary DEPI sesison.
•
The CMTS Core MUST be compliant with [R-DEPI] and [R-UEPI].
•
The RPD MUST be compliant with [R-DEPI] and [R-UEPI].
•
The CMTS Core MUST support the R-DEPI MPT pseudowire type for DOCSIS 3.0 and MPEG-TS Video.
•
The CMTS Core SHOULD support the R-DEPI MCM pseudowire type for DOCSIS 3.0 and MPEG-TS Video.
•
The CMTS Core MUST support all R-DEPI PSP pseudowire types for DOCSIS 3.1.
•
The CMTS Core MUST support all R-UEPI PSP pseudowire types.
•
The RPD MUST support the R-DEPI MPT pseudowire type for DOCSIS 3.0 and MPEG-TS Video.
•
The RPD MUST support the R-DEPI MCM pseudowire type for DOCSIS 3.0 and MPEG-TS Video.
•
The RPD MUST support all R-DEPI PSP pseudowire types for DOCSIS 3.1.
•
The RPD MUST support all R-UEPI PSP pseudowire types.
06/15/15
CableLabs®
47
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
7 SECURE SOFTWARE DOWNLOAD
7.1
Introduction
Remote PHY architecture supports downloading code to RPDs. Authenticating the source and verifying the integrity
of downloaded code is vital to the overall operation and security of Remote PHY architecture. The methods for secure
software download as well as the relevant specification text have been adopted from DOCSIS 3.1 Security
Specification [SECv3.1].
Broadly speaking, with respect to secure software downloads; the RPD assumes the functions of a DOCSIS cable
modem. It is envisioned that such an approach will allow the operators to reuse the majority of the OSS infrastructure
deployed for CM software and security certificate management to perform equivalent functions for RPDs. However,
there are important changes to the upgrade procedure. These changes are summarized below and explained further
within section 7.
•
The RPD upgrade process relies on certificates from the new CableLabs PKI. The legacy certificates are not
supported.
•
Unlike a DOCSIS CM, the RPD does not receive a configuration file from a provisioning system. RPD
initialization involves connecting to and obtaining configuration information from a Principal CCAP-Core via
GCP. The software upgrade TLVs received via GCP effectively replace equivalent TLVs received by a CM in a
configuration file.
•
Unlike a CM SSU process, which needs to be enabled by inclusion of CVC in the CM configuration file, the RPD
is implicitly enabled for SSU, unless the Principal CCAP-Core disables this feature.
•
A DOCSIS CM receives time service from the provisioning system via the DOCSIS Time Protocol, while an
active RPD is time synchronized via the PTP protocol. The RPD needs to receive time service via the DOCSIS
Time Protocol before it establishes the PTP protocol connection.
The RPD code is signed with a certificate from the new PKI defined in [SECv3.1] and then validated by the RPD. The
software download module is an attractive target for an attacker. If an attacker were able to mount an attack against the
software download module, s/he could potentially install code to disrupt service on a wide scale or to redirect the
content. To thwart these attacks, the attacker is forced to overcome several security barriers.
7.2
Overview
The requirements defined in this section address these security objectives for the code download process:
•
The RPD needs to have a means to authenticate that the originator of any download code is a known and trusted
source;
•
The RPD needs to have a means to verify that the downloaded code has not been altered from the original form in
which it was provided by the trusted source;
•
The process needs to simplify the operator’s code file-handling requirements and provide mechanisms for the
operator to upgrade or downgrade the code version of RPDs on their network;
•
The process allows operators to dictate and control their policies with respect to: 1) which code files will be
accepted by RPDs within their network; and 2) security controls that establish the security of the process on their
network;
•
RPDs are able to move freely among systems controlled by different operators;
•
Support updating the Root CA Certificate in the RPD (optional);
•
Support updating the Device CA Certificate in the RPD (optional).
The concerns of individual operators or RPD manufacturers may result in additional security related to the distribution
or installation of code into a RPD. This specification does not restrict the use of further protections, as long as they do
not conflict with the requirements of this specification.
48
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
Multiple levels of protection are required to protect and verify the code download:
•
The manufacturer of the RPD code always applies a digital signature to the code file. The signature is verified
with a certificate chain that extends up to the Root CA before accepting a code file. The manufacturer signature
affirms the source and integrity of the code file to the RPD;
•
Though the manufacturer always signs its code file, an operator may later apply its code signature in addition to
the manufacturer signature. If a second signature is present, the RPD verifies both signatures with a certificate
chain that extends up to the Root CA before accepting a code file;
•
OSS mechanisms for the provisioning and control of the RPD are critical to the proper execution of this process.
Code downloads are initiated by the Principal CCAP-Core during the initial RPD configuration process, or can be
initiated in normal operation using an SNMP command.
The RPD code file is built using a [PKCS#7]-compliant structure that is defined below, which is identical to the code
structure used to upgrade CM software. Included in this structure are:
•
The upgrade code image;
•
The Code Verification Signature (CVS); i.e., the digital signature over the code image and any other authenticated
attributes as defined in the structure;
•
The Code Verification Certificate (CVC); i.e., an [X.509]-compliant certificate that is used to deliver and validate
the public code verification key that will verify the signature over the code image. The DOCSIS Certificate
Authority (CA), a trusted party whose public key is already stored in the RPD, signs this certificate.
Figure 18 shows the basic steps required for the signing of a code image when the code file is signed only by the RPD
manufacturer, and when the code file is signed by the RPD manufacturer and co-signed by an operator.
In DOCSIS, the Root CA certificate is installed in each RPD as a trust anchor. The code manufacturer builds the code
file by signing the code image using a DOCSIS [PKCS#7] digital signature structure with a Manufacturer CVC
certificate and the issuing CVC CA certificate. The code file is then sent to the operator. The operator verifies that the
code file is from a trusted DOCSIS manufacturer and has not been modified. At this point, the operator has the option
of loading the code file on the Software Download server as-is, or of adding its signature and operator CVC and
issuing CVC CA certificate to the code file. During the code upgrade process, the RPD retrieves the code file from the
Software Download server and verifies the new code image using the Root CA Certificate trust anchor before
installing it. See Annex E for CVC chain details.
06/15/15
CableLabs®
49
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
CVC CA
Signed Code Verification Certificate
Code Manufacturer(s)
Build & Test Code File
Sign Code
Send Mfg-signed
Code File (includes
CVS and CVC) to
System Operator
Root Public Key
Root Public Key
System
Operator(s)
Verify Code File
(recommended)
Add MSO Co -Signature and
MSO CVC to code file
Send Code File to RPD upon
request
Root Public
Key
System
Operator(s)
Verify Code File
(recommended)
Send Code File to
RPD upon request
Root Public Key
RPD(s)
RPD(s)
Download Code File
Verify Mfg-signed/
MSO co-signed
Code File
If Code File verifies,
install code image
Download Code File
Verify Mfg-signed
Code File
If Code File verifies,
install code image
Manufacturer – Only Signed
Code Files
Manufacturer – Signed and
MSO Co-Signed Code Files
Figure 18 - Typical Code Validation Hierarchy
7.3
RPD Software Upgrade Procedure
Once the RPD is authenticated (or bypasses authentication) and has obtained an IP address, it may receive a software
update. The software of the RPD may occur by one of the following methods:
1.
During RPD operational configuration from the Principal CCAP-Core or at any time when operational, the
software may be updated using the GCP software update option.
2.
An external management system can connect to the R-PHY and command it to update its software via SNMP.
This option is available even before the RPD is connected to any CCAP device.
The RPD MUST support a software update initiated through the GCP software update feature. The RPD MUST
support a software update initiated via SNMP.
50
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
A software update is accomplished by providing the RPD a filename, an IP address (v4 or v6) for a software download
server and a Code Validation Certificate (CVC). The RPD then uses TFTP or HTTP to go to that server to obtain the
software update.
NOTE: This method of software update is intentionally similar to how a DOCSIS CM is assigned a new software image
so that the existing DOCSIS infrastructure may be leveraged.
The RPD MUST implement a TFTP client compliant with [RFC 1350] for software file downloads. The RPD MAY
implement an HTTP-client compliant with [RFC 1945] or [RFC 2616] for software file downloads. The transfer is
SNMP-initiated, as described in [CCAP-OSSI v3.1], or initiated by the CCAP-Core via GCP, as described here.
The RPD MUST include the TFTP block size option [RFC 2348] when requesting the software image file.
The RPD MUST request a block size of 1448 octets if using TFTP over IPv4.
The RPD MUST request a block size of 1428 octets if using TFTP over IPv6.
If the file specified in the GCP Software Upgrade File Name TLV does not match the current software image of the
RPD, the RPD MUST request the specified file via TFTP from the software server. The RPD selects the software
download server as follows:
•
If the RPD communicates with CCAP-Core via IPv4 and receives the Software Upgrade IPv4 TFTP Server
TLV via GCP, the RPD MUST use the server specified by this TLV. The RPD MUST ignore the Software
Upgrade IPv6 TFTP Server TLV when it communicates with CCAP-Core via using IPv4.
•
If the RPD communicates with CCAP-Core via IPv6 and receives the Software Upgrade IPv6 TFTP Server
TLV via GCP, the RPD MUST use the server specified by this TLV. The RPD MUST ignore the Software
Upgrade IPv4 TFTP Server TLV when it communicates with CCAP-Core via IPv6.
When performing a GCP-initiated software download, the RPD MAY defer normal operation until the download is
complete. The RPD MUST verify that the downloaded image is appropriate for itself. If the image is appropriate, the
RPD MUST write the new software image to non-volatile storage. Once the file transfer is completed successfully,
and the software image is verified, the RPD MUST restart itself with the new code image with a RPD Initialization
Reason of SW_UPGRADE_REBOOT.
If the RPD is unable to complete the file transfer for any reason, it MUST remain capable of accepting new software
download requests (without operator or user interaction), even if power or connectivity is interrupted between
attempts. The RPD MUST log the failure. The RPD MAY report the failure asynchronously to the network
manager. The RPD MUST continue to operate with the existing software if an upgrade cannot be performed.
If the RPD receives a valid image, it will automatically upgrade its software, reboot and repeat the entire initialization
process, including authentication. Image validation uses the same method involving digital signatures and the PKI
certificate as defined in the DOCSIS secure software download process.
If a Principal CCAP-Core initiates a Remote PHY software upgrade during operational configuration or at any time
during active operation, then the following events occur:
•
The GCP session to the Principal CCAP-Core initiating the update is terminated;
•
Any L2TPv3 connections to the Principal CCAP-Core initiating the update are terminated;
•
Any active GCP and L2TPv3 connections to other CCAP-Cores are terminated;
•
The software upgrade is performed;
•
The RPD reboots.
If a network management entity initiates the software upgrade, then the following events occur:
•
Any active GCP and L2TPv3 connections to all CCAP-Cores are terminated;
•
The software upgrade is performed;
•
The RPD reboots.
06/15/15
CableLabs®
51
CM-SP-R-PHY-I01-150615
7.4
Data-Over-Cable Service Interface Specifications
Software Code Upgrade Requirements
The following sections define the requirements of the RPD software code upgrade verification process. All RPD code
upgrades are prepared and verified as described. All RPDs MUST verify code upgrades according to this
specification. The new PKI used for issuing CVCs consists of three types of certificates: a Root CA, a CVC CA, and
the CVC. CableLabs manages the new PKI and the certificates issued from its CAs (CableLabs Root CA and
CableLabs CVC CA); see [SECv3.1] for certificate profile and extension definitions. The RPD MUST process CVC
extensions as defined by [RFC 5280]
NOTE: The CableLabs Root CA is used to issue both RPD Device Certificates and CVC Certificates. RPDs do not
support the code upgrade requirements that use the legacy PKI defined in DOCSIS 3.0.
7.4.1
Code File Processing Requirements
The code file format is defined in the [SECv3.1].
The RPD MUST reject the DOCSIS [PKCS#7] code file if the signedData field does not match the DER-encoded
structure represented in [SECv3.1].
The RPD MUST be able to verify DOCSIS code file signatures that are signed using key modulus lengths of 1024,
1536, and 2048 bits. The public exponent is F4 (65537 decimal).
The RPD MUST reject the CVC if it does not match the DER-encoded structure represented in [SECv3.1].
The RPD MUST NOT install the upgraded code image unless the code image has been verified as being compatible
with the RPD.
If the code download and installation is successful, then the RPD MUST replace its currently stored Root CA
Certificate with the Root CA Certificate in the SignedContent field, if one was present.
If the code download and installation is successful, then the RPD MUST replace its currently stored Device CA
Certificate with the Device CA Certificate received in the SignedContent field, if any were present.
7.4.2
Code File Access Controls
In addition to the cryptographic controls provided by the digital signature and the certificate, special control values are
included in the code file for the RPD to check before it accepts a code image as valid. The conditions placed on the
values of these control parameters MUST be satisfied before the RPD attempts to validate the CVC and the CVS (see
Sections 7.4.3.1 and 7.4.3.2).
7.4.2.1
Subject Organization Names
The RPD MUST recognize up to two names that it considers a trusted code-signing agent if present in the subject field
of a code file CVC. These are:
•
The RPD manufacturer: The RPD MUST verify that the manufacturer name in the manufacturer CVC subject
field exactly matches the manufacturer name stored in the RPD’s non-volatile memory by the manufacturer. A
manufacturer CVC is always included in the code file.
•
A co-signing agent: DOCSIS technology permits another trusted organization to co-sign code files destined for
the RPD. In most cases this organization is the operator. The organization name of the co-signing agent is
communicated to the RPD via a co-signer CVC via GCP when initializing the RPD’s code verification process.
The RPD MUST verify that the co-signer organization name in the co-signer CVC subject field exactly matches
the co-signer organization name previously received in the co-signer initialization CVC, and stored by the RPD.
7.4.2.2
Time Varying Controls
In support of the code upgrade process, the RPD MUST keep two UTC time values associated with each code-signing
agent. These values are known as codeAccessStart and cvcAccessStart. The RPD MUST store and maintain one
pair of time values for the RPD manufacturer signing agent. If the RPD is assigned a code co-signing agent, the RPD
MUST maintain a pair of time values for the code co-signing agent.
52
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
These values are used to control code file access to the RPD by individually controlling the validity of the CVS and the
CVC. Stored and maintained time values in the RPD MUST have a precision of one second. Stored and maintained
time values in the RPD MUST be capable of representing all times (with one second precision) between midnight,
January 1 1950 and midnight January 1 2050.
The RPD MUST NOT allow the values of codeAccessStart and cvcAccessStart corresponding to the RPD’s
manufacturer signing agent to decrease. The RPD MUST NOT allow the value of codeAccessStart and
cvcAccessStart corresponding to the co-signing agent to decrease as long as the co-signing agent does not change
and the RPD maintains co-signer time-varying control values (see Section 7.4.5).
7.4.3
RPD Code Upgrade Initialization
Before the RPD can upgrade code, it should be properly initialized. The manufacturer first initializes the RPD.
7.4.3.1
Manufacturer Initialization
It is the responsibility of the manufacturer to install the initial code version in the RPD.
In support of code upgrade verification, values for the following parameters MUST be loaded into the RPD’s
non-volatile memory:
•
RPD manufacturer organizationName;
•
codeAccessStart initialization value;
•
cvcAccessStart initialization value.
The RPD MUST initialize the values of codeAccessStart and cvcAccessStart to an UTCTime equal to the validity
start time of the manufacturer’s latest CVC. These values will be updated periodically under normal operation via
manufacturer CVCs that are received and verified by the RPD.
7.4.3.2
Operational Initialization
The method for obtaining RPD code download files is defined in Section 7.4.3. The RPD receives settings relevant to
code upgrade verification from the Principal CCAP-Core via GCP. The RPD MUST NOT use these settings until after
the Principal CCAP-Core has successfully completed the operational bringup of the RPD.
The GCP TLVs normally include the most up-to-date CVC applicable for the destination RPD. When the CCAP-Core
initiates a code upgrade, it provides a CVC to initialize the RPD for accepting code files according to this
specification. Regardless of whether a code upgrade is required, a CVC in the GCP TLVs MUST be processed by the
RPD.
The Principal CCAP-Core enables or disables the RPD’s capability to initiate SSU from SNMP through “SSU
Control” TLV. This mechanism can be utilized to prevent software upgrades from SNMP interrupting vital services.
The Principal CCAP-Core can, for example, restrict the time interval in which software upgrades can be performed to
coincide with the MSO’s service window.
The RPD is effectively enabled for SSU from SNMP unless it is connected to a Principal CCAP-Core and the
Principal CCAP-Core explicitly disables SSU. When the SSU is disabled via GCP, the RPD MUST reject any attempt
to initiate SSU from SNMP.
After the RPD is disconnected from the Principal CCAP-Core, the SSU upgrade capability is effectively enabled
regardless of the settings established earlier by the Principal CCAP-Core.
GCP TLVs may contain:
•
A “SSU Control” TLV disabling or enabling the SW upgrade capability.
•
No CVCs;
•
From DOCSIS 3.1 PKI:
•
A Manufacturer CVC Chain (the Manufacturer CVC and its issuing CA certificate);
•
A Co-signer CVC Chain (the Co-signer CVC and its issuing CA certificate);
06/15/15
CableLabs®
53
CM-SP-R-PHY-I01-150615
•
Data-Over-Cable Service Interface Specifications
Both Manufacturer CVC Chain and Co-signer CVC Chain
When the RPD has not received a co-signer CVC, the RPD MUST NOT accept code files that have been co-signed.
If the RPD is configured to accept code co-signed by a code-signing agent, the following parameters MUST be stored
in the RPD’s memory when the co-signer CVC is processed:
•
Co-signing agent’s organizationName;
•
Co-signer cvcAccessStart;
•
Co-signer codeAccessStart.
Unlike the manufacturer organizationName and time varying control values, the co-signer organizationName
and time varying control values are not required to be stored in non-volatile memory.
7.4.3.2.1
Processing the CVC Received via GCP
When a CVC is included in the GCP TLVs, the RPD MUST verify the CVC before accepting any of the code upgrade
settings it contains. Upon receipt of the CVC the RPD MUST perform the following validation and procedural steps.
•
If any of the following verification checks fail, the RPD MUST immediately halt the CVC verification process.
•
If the GCP TLVs do not include a valid CVC, the RPD MUST NOT download upgrade code files, triggered by
the GCP.
•
If the GCP TLVs do not include a valid CVC, the RPD MUST NOT accept information from a CVC subsequently
delivered via an SNMP MIB.
Following receipt of a CVC via GCP, and after the RPD has successfully became operational with the Principal
CCAP-Core, the RPD MUST:
1.
Verify that the Extended Key Usage extension is present in the CVC, as specified in Appendix III of [SECv3.1].
2.
Verify that the manufacturer CVC validity start time is greater than or equal to the manufacturer
cvcAccessStart value currently held in the RPD if the CVC is a Manufacturer CVC and the subject
organizationName is identical to the RPD's manufacturer name.
3.
Reject this CVC and log an error if the CVC is a Manufacturer CVC and the subject organizationName is not
identical to the RPD’s manufacturer name.
4.
Verify that the validity start time is greater than or equal to the co-signer cvcAccessStart value currently held in
the RPD if the CVC is a Co-signer CVC and the subject organizationName is identical to the RPD’s current
code co-signing agent.
5.
After the CVC has been validated, make this subject organization name become the RPD’s new code co-signing
agent if the CVC is a Co-signer CVC and the subject organizationName is not identical to the current code
co-signing agent name.
6.
Verify that the CVC and any CVC CA Certificate signatures chain up to the Root CA Certificate of the new PKI
held by the RPD.
7.
Verify that the validity periods for the CVC and the issuing CA certificate have not expired.
8.
Update the RPD's current value of cvcAccessStart corresponding to the CVC’s subject organizationName
(i.e., manufacturer or code co-signing agent) with the validity start time value from the validated CVC. If the
validity start time value is greater than the RPD’s current value of codeAccessStart, update the RPD’s
codeAccessStart value with the validity start time value.
7.4.3.2.2
Processing the SNMP CVC
The RPD MUST process CVCs received via SNMP when it is enabled to perform SSU. When the RPD is disabled
from performing SSU it MUST reject all CVCs received via SNMP. CVCs received via SNMP will also chain up to
the same Root CA certificate or public key that was used to validate the CVC from GCP (see Section 7.4.5). When
validating a CVC received via SNMP, the RPD MUST perform the following validation and procedural steps. If any
54
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
of the following verification checks fail, the RPD MUST immediately halt the CVC verification process, log the error
if applicable, and remove all remnants of the process up to that step.
When a RPD receives a CVC via SNMP, it MUST:
1.
Verify that the Extended Key Usage extension is in the CVC as specified in Appendix III of [SECv3.1].
2.
Verify that the manufacturer CVC validity start time is greater than the manufacturer cvcAccessStart value
currently held in the RPD if the CVC subject organizationName is identical to the RPD’s manufacturer name.
3.
Verify that the validity start time is greater than the co-signer cvcAccessStart value currently held in the RPD if
the CVC subject organizationName is identical to the RPD’s current code co-signing agent.
4.
Reject this CVC if the CVC subject organizationName is not identical to RPD’s manufacturer or current code
co-signing agent name.
5.
Verify that the CVC and any CVC CA Certificate signatures chain up to the same Root CA Certificate or Root CA
key that was used to validate the corresponding CVC (manufacturer or co-signer) from GCP.
6.
Verify that the validity periods for the CVC and the issuing CA certificate have not expired.
7.
Update the current value of the subject's cvcAccessStart values with the validated CVC’s validity start time
value. If the validity start time value is greater than the RPD’s current value of codeAccessStart, the RPD
MUST replace its codeAccessStart value with the validity start value.
7.4.4
Code Signing Guidelines
Manufacturer and operator code signing guidelines are provided in Appendix III of [SECv3.1].
7.4.5
Code Verification Requirements
The RPD MUST NOT install upgraded code unless the code has been verified.
7.4.5.1
RPD Code Verification Steps
When downloading code, the RPD MUST perform the verification checks presented in this section. If any of the
verification checks fail, or if any section of the code file is rejected due to invalid formatting, the RPD MUST
immediately halt the download process and log the error if applicable, remove all remnants of the process to that step,
and continue to operate with its existing code. The verification checks can be made in any order.
1.
The RPD MUST verify that:
•
2.
The value of signingTime is equal to or greater than the manufacturer codeAccessStart value currently
held in the RPD;
• The value of signingTime is equal to or greater than the manufacturer CVC validity start time;
• The value of signingTime is less than or equal to the manufacturer CVC validity end time.
The RPD MUST verify that:
•
3.
The manufacturer CVC subject organizationName is identical to the manufacturer name currently stored in
the RPD’s memory;
• The manufacturer CVC validity start time is equal to or greater than the manufacturer cvcAccessStart
value currently held in the RPD;
• The Extended Key Usage extension in the Manufacturer CVC meets the requirements of Appendix III of
[SECv3.1];
The RPD MUST verify that the Mfr CVC chains up to the Root CA held by the RPD.
4.
Verify that the validity periods for the CVC and the issuing CA certificate have not expired.
5.
The RPD MUST verify the manufacturer code file signature. If the signature does not verify, the RPD MUST
reject all components of the code file (including the code image), and any values derived from the verification
process should be immediately discarded.
06/15/15
CableLabs®
55
CM-SP-R-PHY-I01-150615
6.
Data-Over-Cable Service Interface Specifications
If the manufacturer signature verifies and a co-signing agent signature is required:
a)
The RPD MUST verify that:
(1) The co-signer signature information is included in the code file;
(2) The value of signingTime is equal to or greater than the corresponding codeAccessStart value
currently held in the RPD;
(3) The value of signingTime is equal to or greater than the corresponding CVC validity start time;
(4) The value of signingTime is less than or equal to the corresponding CVC validity end time.
b) The RPD MUST verify that:
(1) The co-signer CVC subject organizationName is identical to the co-signer organization name currently
stored in the RPD’s memory;
(2) The co-signer CVC validity start time is equal to or greater than the cvcAccessStart value currently
held in the RPD for the corresponding subject organizationName;
(3) The Extended Key Usage extension in the Co-signer CVC meets the requirements of Appendix III of
[SECv3.1].
c) The RPD MUST verify that the Co-Signing CVC Certificate chains up to the Root CA held by the RPD.
d) The RPD MUST verify that the validity periods for the CVC and the issuing CA certificate have not expired.
e) The RPD MUST verify the co-signer code file signature. If the signature does not verify, the RPD MUST
reject all components of the code file (including the code image), and any values derived from the verification
process should be immediately discarded.
7.
Once the manufacturer, and optionally the co-signer, signature has been verified, the code image can be trusted
and installation may proceed. Before installing the code image, all other components of the code file and any
values derived from the verification process except the [PKCS#7] signingTime values and the CVC validity start
values SHOULD be immediately discarded.
8.
The RPD upgrades its software by installing the code file according to Section 7.3.
9.
If the code installation is unsuccessful, the RPD MUST discard the [PKCS#7] signingTime values and CVC
validity start values it just received in the code file. The procedure for handling this failure condition is specified
in [MULPI v3.1].
10. Once the code installation is successful, the RPD MUST:
a) Update the current value of manufacturer codeAccessStart with the [PKCS#7] signingTime value;
b) Update the current value of manufacturer cvcAccessStart with the CVC validity start value.
11. If the code installation is successful, and if the code file was co-signed, the RPD MUST:
a) Update the current value of the co-signer codeAccessStart with the [PKCS#7] signingTime value;
b) Update the current value of the co-signer cvcAccessStart with the CVC validity start value.
7.4.6
DOCSIS Interoperability
Images for RPD secure software download are to be signed using certificates from the new PKI defined in the
[SECv3.1] specification. Images for legacy secure software download are signed using certificates from the legacy
PKI defined in [SECv3.0]are not supported by RPDs. The RPD supports secure software downloads using certificates
only from the new PKI.
7.4.7
Error Codes
The RPD MUST log the following error events when they occur during the code verification process. RPD event
logging requirements and event message format are defined in [R-OSSI].
1.
Improper code file controls
Conditions:
a)
56
CVC subject organizationName for manufacturer does not match the RPD’s manufacturer name.
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
b) CVC subject organizationName for code co-signing agent does not match the RPD’s current code
co-signing agent.
c) The manufacturer [PKCS#7] signingTime value is less-than the codeAccessStart value currently held in
the RPD.
d) The manufacturer [PKCS#7] validity start time value is less-than the cvcAccessStart value currently held
in the RPD.
e) The manufacturer CVC validity start time is less-than the cvcAccessStart value currently held in the RPD.
f) The manufacturer [PKCS#7] signingTime value is less-than the CVC validity start time.
g) Missing or improper extended key-usage extension in the manufacturer CVC.
h) The co-signer [PKCS#7] signingTime value is less-than the codeAccessStart value currently held in the
RPD.
i) The co-signer [PKCS#7] validity start time value is less-than the cvcAccessStart value currently held in
the RPD.
j) The co-signer CVC validity start time is less-than the cvcAccessStart value currently held in the RPD.
k) The co-signer [PKCS#7] signingTime value is less-than the CVC validity start time.
l) Missing or improper extended key-usage extension in the co-signer CVC.
2.
Code file manufacturer CVC validation failure
Conditions:
a)
The manufacturer CVC in the code file does not chain to the same root CA as the manufacturer CVC received
via GCP.
3.
Code file manufacturer CVS validation failure
4.
Code file co-signer CVC validation failure
Conditions:
a)
The co-signer CVC in the code file does not chain to the same root CA as the co-signer CVC received via
GCP.
5.
Code file co-signer CVS validation failure.
6.
Improper format of CVC received via GCP.
Conditions:
a)
Missing or improper key usage attribute.
7.
Validation failure of CVC received via GCP.
8.
Improper SNMP CVC format.
Conditions:
a) CVC subject organizationName for manufacturer does not match the RPD’s manufacturer name.
b) CVC subject organizationName for code co-signing agent does not match the RPD’s current code
co-signing agent.
c) The CVC validity start time is less-than or equal-to the corresponding subject's cvcAccessStart value
currently held in the RPD.
d) Missing or improper key usage attribute.
9.
SNMP CVC validation failure.
Conditions:
a)
The manufacturer CVC received via SNMP does not chain to the same root CA as the manufacturer CVC
received via GCP.
b) The co-signer CVC received via SNMP does not chain to the same root CA as the co-signer CVC in the
received via GCP.
06/15/15
CableLabs®
57
CM-SP-R-PHY-I01-150615
7.5
Data-Over-Cable Service Interface Specifications
Security Considerations (Informative)
The method(s) used to protect private keys are a critical factor in maintaining security. Users authorized to sign code,
i.e., manufacturers and operators who have been issued code verification certificates (CVCs) by the DOCSIS root CA,
should protect their private keys. An attacker with access to the private key of an authorized code-signing user can
create, at will, code files that are potentially acceptable to a large number of RPDs.
The defense against such an attack is for the operator to revoke the certificate whose associated code-signing private
key has been learned by the attacker. To revoke a certificate, the operator delivers to each affected RPD, an updated
CVC with a validity start time that is newer than that of the certificate(s) being revoked. The new CVC can be
delivered via any of the supported mechanisms: GCP, code file, or SNMP. The new CVC implicitly revokes all
certificates whose validity start time is earlier than that of the new CVC.
To reduce the vulnerability to this attack, operators should regularly update the CVC in each RPD, at a frequency
comparable to how often the operator would update a CRL if one were available. Regular updates help manage the
time interval during which a compromised code-signing key is useful to an attacker. CVCs should also be updated if it
is suspected that a code-signing key has been compromised. To update the CVC, the user needs a CVC whose validity
start time is newer than the CVC in the RPD. This implies that the DOCSIS root CA regularly issues new CVCs to all
authorized code-signing manufacturers and operators, to make the CVCs available for update.
When an RPD is attempting to become operational with the Principal CCAP-Core for the first time or after being
off-line for an extended period, it should receive a trusted CVC as soon as possible. This provides the RPD with the
opportunity to receive the most up-to-date CVC available and deny access to CVCs that needed to be revoked since the
RPD’s last initialization. The first opportunity for the RPD to receive a trusted CVC is via GCP from the Principal
CCAP-Core. The Principal Core has the ability to control RPD’s ability to perform the SSU through a dedicated GCP
TLV. If the Principal Core disables SSU, the RPD will not request or have the ability to remotely upgrade code files. In
addition, the RPD will not accept CVCs subsequently delivered via SNMP.
To mitigate the possibility of an RPD receiving a previous code file via a replay attack, the code files include a
signing-time value in the [PKCS#7] structure that can be used to indicate the time the code image was signed. When
the RPD receives a code file signing-time that is later than the signing-time it last received, it will update its internal
memory with this value. The RPD will not accept code files with an earlier signing-time than this internally stored
value. To upgrade an RPD with a new code file without denying access to past code files, the signer may choose not to
update the signing-time. In this manner, multiple code files with the same code signing-time allow an operator to
freely downgrade an RPD’s code image to a past version (that is, until the CVC is updated). This has a number of
advantages for the operator, but these advantages should be weighed against the possibilities of a code file replay
attack.
Without a reliable mechanism to revert back to a known good version of code, any code-update scheme, including the
one in this specification, has the weakness that a single, successful forced update of an invalid code image may render
the RPD useless, or may cause the RPD to behave in a manner harmful to the network. Such an RPD may not be
repairable via a remote code update, since the invalid code image may not support the update scheme.
58
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
8 X.509 CERTIFICATE PROFILE AND MANAGEMENT
R-PHY employs X.509 version 3 digital certificates for authenticating key exchanges between RPD and NAD and
between RPD and CCAP-Core. [X.509] is a general-purpose standard; the certificate profile, described here, further
specifies the contents of the certificate’s defined fields. This certificate profile also defines the hierarchy of trust for
the management and validation of certificates.
Except where otherwise noted in Annex E, the certificates used comply with [RFC 5280].
8.1
Certificate Management Architecture Overview
The certificate management architecture for RPD authentication uses the DOCSIS 3.1 PKI defined by [SECv3.1]. The
PKI consists of a three-level hierarchy of trust supporting three types of certificates:
•
Root CA Certificate;
•
Device CA Certificate;
•
RPD Device Certificates.
The Root CA Certificate is used as a trust anchor for the PKI and issues the Device CA Certificate that issues the RPD
Device Certificates. The PKI uses a "centralized" model where the Device CA is hosted by CableLabs or an approved
3rd party that issues RPD Device Certificates to approved manufacturers. CableLabs manages the PKI and the
certificates issued from its CAs (for information about CableLabs Root CA and CableLabs Device CA, see Annex E).
The Root CA will also be used as a trust anchor for issuing and validating CA and Code Verification Certificates
(CVCs) for the Secure Software Download (SSD) process specified in Section 7.
The Root CA generates and distributes to operators a Certificate Revocation List (CRL), identifying revoked
manufacturer certificates. The manner in which CRLs are distributed is outside the scope of this specification. In order
to reduce the burden on RPD devices that are designed to work in multiple geographic regions, an effort will be made
to consolidate the DOCSIS 3.1 PKI hierarchy such that the same device certificate for DOCSIS 3.1 will also be valid
for EuroDOCSIS 3.1 and other international versions of DOCSIS 3.1 and above.
8.2
RPD Certificate Storage and Management in the RPD
The RPD MUST have a factory installed RPD Device Certificate (and associated private keys) that is issued from the
new PKI. The RPD uses the RPD Device Certificate when authenticating with a NAD or CCAP-Core.
The RPD’s non-volatile memory MUST contain a Root CA certificate for SSD image verification.
The RPD MAY be capable of updating or replacing the Device CA Certificate via the DOCSIS code download file
(see Section 7).
The RPD MUST be able to process certificate serial number values containing 20 octets or fewer. The RPD MUST
accept certificates that have serial numbers that are negative or zero.
8.3
Certificate Processing and Management in the CCAP-Core
IKEv2 (see [RFC 7296]) employs digital certificates to verify the binding between a device’s identity (encoded in a
digital certificate’s subject name) and its public key. The CCAP-Core does this by validating the RPD Device
Certificate’s certification path. This path will typically consist of three chained certificates: the RPD Device
Certificate, the Device CA certificate and the Root CA certificate (see section 8.1). Validating the chain follows the
"Basic Path Validation" rules defined in [RFC 5280].
The CCAP-Core MUST support validating certificate chains from the DOCSIS 3.1 PKI.
[RFC 4131] requires that CCAP-Cores support administrative controls that allow the operator to override certification
chain validation by identifying a particular CA or RPD Device Certificate as trusted or untrusted. This section
specifies the management model for the exercise of these controls, as well as the processing a CCAP-Core undertakes
06/15/15
CableLabs®
59
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
to assess a RPD Device Certificate’s validity, and thus verify the binding between the RPD’s identity and its public
key.
The CCAP-Core MUST be able to process certificate serial number values containing 20 octets or fewer. The
CCAP-Core MUST accept certificates that have serial numbers that are negative or zero.
Annex E describes the format of the subject name field for each type of DOCSIS certificate. The issuer field of a
certificate exactly matches the subject field of the issuing certificate. DOCSIS 3.1 PKI certificates transmitted by an
RPD have name fields that conform to the format described in Annex E. A CCAP-Core MUST be capable of
processing the name fields of a certificate if the name fields conform to the indicated format in Annex E. A
CCAP-Core MAY choose to accept a certificate that has name fields that do not conform to the indicated format in
Annex E.
The CCAP-Core MUST process certificate extensions as defined by [RFC 5280] (see Annex E for certificate profile
and extension definitions).
8.3.1
CCAP-Core Certificate Management Model
The CCAP-Core holds copies of the Root CA, Device CA, and RPD Device Certificates (see Section 8.1), which it
obtains in one of two ways: 1) provisioning, or 2) IKEv2 messaging. Each certificate learned by a CCAP-Core MUST
be assigned one of four states:
•
Untrusted,
•
Trusted,
•
Chained, or
•
Root.
The CCAP-Core MUST support the ability to provision at least two Root CA Certificates. The CCAP-Core MUST
support the ability to display the entire Root Certificate(s) and/or its thumbprint to the operator.
A CCAP-Core learns of Device CA certificates through either the CCAP-Core’s provisioning interface or through
receipt and processing of the client RPDs’ Authentication Information messages. Regardless of how a CCAP-Core
obtains its Device CA certificates, the CCAP-Core MUST mark them as either Untrusted, Trusted, or Chained. If a CA
Certificate is not self-signed, the CCAP-Core MUST mark the certificate as Chained. The CCAP-Core, however,
MUST support administrative controls that allow an operator to override the Chained marking and identify a given CA
certificate as Trusted or Untrusted.
If a Device CA Certificate is self-signed, the CCAP-Core MUST mark the certificate as either Trusted or Untrusted,
according to administratively controlled CCAP-Core policy.
A CCAP-Core obtains copies of RPD Device Certificates in the IKEv2 messages it receives from RPDs. RPD Device
Certificates are issued by a Device CA. Thus, the CCAP-Core MUST mark RPD Device Certificates as Chained
unless overridden by CCAP-Core administrative control and configured as Trusted or Untrusted.
8.3.2
Certificate Validation
The CCAP-Core validates the certification paths of CA and RPD Device Certificates using Basic Path Validation rules
defined in [RFC 5280] and the criteria below.
The CCAP-Core MUST label CA and RPD Certificates as Valid or Invalid if their certification paths are valid or
invalid respectively. Trusted certificates, provisioned in the CCAP-Core, MUST be Valid; this is true even if the
current time does not fall within the Trusted certificate's validity period. Untrusted certificates, provisioned in the
CCAP-Core, MUST be Invalid.
The CCAP-Core MUST mark a chained certificate as Valid only if:
1.
The certificate chains to a Root CA, Trusted, or Valid certificate that has not been revoked as defined by the Basic
Path Validation section in [RFC 5280]; and
2.
The current time falls within the validity period of each Chained or Root certificate within the certificate chain;
and
60
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
3.
The certificate is not identified as revoked (see Section 8.4); and
4.
In the case of a RPD Device Certificate, the RPD MAC address encoded in its tbsCertificate.subject field
and RSA public key encoded in its tbsCertificate.subjectPublicKeyInfo field match the RPD MAC
address and RSA public key encoded in the IKEv2 messaging; and
5.
In the case of an RPD Device Certificate, if the KeyUsage extension is present, the digitalSignature and/or
keyAgreement bits are turned on, the keyEncipherment bit is turned on, and the keyCertSign and cRLSign bits are
off. In the case of a Device CA Certificate, if the KeyUsage extension is present, the keyCertSign bit is turned on.
Whether criterion 2 above is ignored MUST be subject to CCAP-Core administrative control.
If validity period checking is enabled and the time of day has not been acquired by the CCAP-Core, a (non-permanent)
authorization reject message MUST be returned by the CCAP-Core in response to an authorization request.
The CCAP-Core MUST NOT invalidate certificates that have non-specified critical extensions (contrary to [RFC
5280]) as long as the certificates satisfy the validity criteria above.
8.4
Certificate Revocation
Providing a mechanism for certificate revocation is a normal part of PKI management. When a certificate is issued, it
is expected to be in use for its entire validity period. However, various circumstances may cause a certificate to
become invalid prior to the expiration of the validity period. Such circumstances include change of name, change of
association between subject and CA, and compromise or suspected compromise of the corresponding private key.
Under such circumstances, the CA needs to revoke the certificate.
Two methods of supporting certificate revocation are defined in this specification: 1) Certificate Revocation Lists
(CRLs), and 2) Online Certificate Status Protocol (OCSP). The CCAP-Core MUST support configuration of none,
one, or both certificate revocation methods to be enabled at the same time.
8.4.1
Certificate Revocation Lists
[RFC 5280] defines a method for revoking certificates using [X.509] Certificate Revocation Lists (CRLs).
Figure 19 shows a framework for managing and distributing CRLs. A CRL is a digitally signed, timestamped list of
certificate serial numbers revoked by a Certificate Authority (CA). When a CA identifies the compromised
certificates, the CA could generate the CRLs itself, or a CA could delegate the CRL generation to a third party CRL
Issuer. The CRL Repository is a system that maintains a database of revoked certificates. A description of the interface
between the CA or CRL Issuer and CRL Repository is outside the scope of this specification.
Figure 19 - CRL Framework
The CCAP-Core retrieves CRL entries from the CRL Repository and uses this information to verify if a certificate
received during the RPD authentication process is revoked.
06/15/15
CableLabs®
61
CM-SP-R-PHY-I01-150615
8.4.1.1
Data-Over-Cable Service Interface Specifications
CCAP-Core CRL Support
The CCAP-Core MUST support retrieval of CRL files formatted as defined in [RFC 5280]. CRL files may identify
revoked certificates that were issued from different CAs. Therefore, the CCAP-Core MUST support extensions related
to indirect CRL files, as defined in [RFC 5280]. The CCAP-Core MUST support HTTP as defined in [RFC 2616] for
downloading CRL files.
Before using the information in a CRL file, the CCAP-Core MUST verify that its digital signature chains to a trusted
root CA. Trusted root CAs are administratively provisioned in the CCAP-Core. If the CRL file digital signature cannot
be verified, the CCAP-Core MUST discard the CRL file. The CCAP-Core MUST validate if a CA certificate or RPD
Device Certificate is revoked during the certificate validation process specified in Section 8.3.2.
If the CRL contains the nextUpdate value, the CCAP-Core MUST refresh the CRL after the specified time has passed.
If the CCAP-Core fails to retrieve the new CRL, it MUST log an event (see [CCAP-OSSI v3.1]) and continue to use its
current CRL. If the CCAP-Core fails to retrieve the new CRL it should attempt to retry retrieval of the CRL file on a
periodic basis. If the CRL does not contain the nextUpdate value, the CCAP-Core MUST refresh the CRL according
to the configured value as defined in [CCAP-OSSI v3.1].
When the CCAP-Core is configured to use a CRL it MUST attempt to retrieve the CRL file each time it starts up.
During CCAP-Core startup it is possible that some RPDs may perform IKEv2 authorization before the CRL file has
been retrieved. When the CCAP-Core is configured to use a CRL and an RPD’s device certificate chain is validated
during CCAP-Core startup before the CRL file is retrieved, the CCAP-Core MUST log an event for that RPD
[CCAP-OSSI v3.1] and bypass CRL checking.
8.4.2
Online Certificate Status Protocol
[RFC 6960] defines an Online Certificate Status Protocol (OCSP) for querying the status of a digital certificate. The
CCAP-Core sends a certificate status request to an OCSP responder when it receives a CA certificate or an RPD
Device Certificate (see Figure 20). The OCSP responder sends a status response indicating that the certificate is either
"good," "revoked," or "unknown." The OCSP responder checks only the revocation status of a certificate; it does not
verify the validity of the certificate itself. The CCAP-Core uses the result from the OCSP responder during the
certificate validation process specified in Section 8.3.2.
Figure 20 - OCSP Framework
The CCAP-Core MUST be capable of acting as an OCSP client as defined in [RFC 6960]. The CCAP-Core SHOULD
cache the OCSP response status for a certificate if the nextUpdate value is present in the OCSP response. If the
CCAP-Core caches the OCSP response status for a given certificate, it MUST retrieve the revocation status from the
cache. Once the nextUpdate time for that certificate has passed, the CCAP-Core MUST continue using the revocation
status value from the cache until an update is retrieved from the OCSP Responder. If the CCAP-Core is unable to
retrieve the OCSP status for an uncached certificate or if the retrieved status is "unknown," the CCAP-Core MUST
log an event [CCAP-OSSI v3.1] and assume the certificate status to be "good."
If the nextUpdate value is not present in the OCSP response, the CCAP-Core MUST NOT cache the OCSP response
status for a certificate. If the CCAP-Core is configured with OCSP Responder information, it MUST send an OCSP
request when a CA certificate or RPD Device Certificate is obtained using the Authentication Information message, or
Authentication Request message respectively, unless there is a valid certificate status in the cache.
When the CCAP-Core is attempting to communicate with the OCSP Responder, the exchange should not significantly
delay the RPD provisioning process. If no response is received, the CCAP-Core MUST proceed using the currently
cached revocation status. For uncached certificate states, the CCAP-Core MUST proceed as if a response with the
status "good" has been received.
62
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
The CCAP-Core MUST support OCSP over HTTP as described in [RFC 6960]. The CCAP-Core MAY generate a
signature in the OCSP request. The CCAP-Core MUST bypass validation of the signature in an OCSP response based
on the configured value as defined in [CCAP-OSSI v3.1].
06/15/15
CableLabs®
63
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
9 PHYSICAL PROTECTION OF KEYS IN THE RPD
RPDs MUST store and maintain the RPD Device Certificate RSA private/public key pairs. The RPD MUST store the
RPD Device Certificate private keys in a manner that deters unauthorized disclosure and modification. Also, RPDs
SHOULD prevent debugger tools from reading the RPD Device Certificate private key in production devices by
restricting or blocking physical access to memory containing this key.
The RPD MUST meet [FIPS 140-2] security requirements for all instances of private and public permanent key
storage.
The RPD MUST meet [FIPS 140-2] Security Level 1. FIPS 140-2 Security Level 1 requires minimal physical
protection through the use of production-grade enclosures. The reader should refer to the cited document for the
formal requirements; however, below is a summary of those requirements.
Under the [FIPS 140-2] classification of "physical embodiments" of cryptographic modules, external RPDs are
"multiple-chip stand-alone cryptographic modules. FIPS 140-2 specifies the following Security level 1 requirements
for multiple-chip stand-alone modules:
•
The chips are to be of production-grade quality, which shall include standard passivation techniques (i.e., a
sealing coat over the chip circuitry to protect it against environmental or other physical damage);
•
The circuitry within the module is to be implemented as a production grade multiple-chip embodiment (i.e., a
printed circuit board, a ceramic substrate, etc.);
•
The module is to be entirely contained within a metal or hard plastic production-grade enclosure, which may
include doors or removable covers.
64
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
10 SYSTEM OPERATION (NORMATIVE)
Once the system is operational, there is very little that happens with the MHAv2 protocols. Most of the operational
features are managed within the DOCSIS protocol that is run transparently over MHAv2.
This section explains some variations to the MHAv2 operational state. One of those variations is the location of the
upstream scheduler.
10.1 DOCSIS Upstream Scheduling
The RPD is intended to be a simple and lightweight extension of the CCAP. MHAv2 permits the upstream scheduler
to be located either centrally or remotely. Note that the [R-UEPI] protocol provides sufficient quality of service
mechanisms that the upstream scheduler can be run centrally.
The advantages of running a centralized upstream scheduler are:
•
Similar CMTS software model to an Integrated CMTS.
•
Similar operational model to an Integrated CMTS.
•
Scheduler software is from the same vendor as the CMTS software.
•
Fewer interoperability problems between different vendors of CCAP-Core and the RPD.
•
Access to Debug mode if there are problems with the remote scheduler.
•
Scalable resources for the scheduler if more CPU power is needed.
The advantages of running a distributed upstream scheduler are:
•
Shorter round-trip delay from request to grant that may impact some aspects of performance.
For plant distances of 100 miles or less, the Remote PHY and I-CMTS systems have nearly identical performance as
the I-CMTS, since the I-CMTS is a centralized scheduler system (because the PHY is also centralized). With the PHY
removed from the CMTS Core, the CMTS Core can be located at distances much greater than the original 100 mile
limit for DOCSIS. In these cases, the REQ-GNT turn around time could be extended by several additional
milliseconds. However, the DOCSIS scheduler is a pipelined system. If the time between grants increases, then the
number of bytes per grant will increase to compensate.
The R-PHY system defaults to a centralized scheduler because the differences in performance are negligible and the
benefits are measureable. Support for a distributed scheduler is not included at this time.
10.1.1 Centralized Scheduling Requirements
The requirements regarding centralized scheduling are:
•
The RPD MUST support operation with a centralized scheduler.
•
The RPD MAY support operation with a distributed scheduler.
10.2 Daisy-chaining of the Backhaul Ethernet Port
The RPD may be located in a node enclosure with other entities that aggregate to the same backhaul link to the CIN.
Two distinct forms of aggregation are supported:
1.
All RPDs connect to an Ethernet switch or hub which then connects to the CIN.
2.
Each RPD is daisy-chained with the next RPD, and the last RPD connects to the CIN. In the case of
daisy-chaining, it is as if each RPD has a three-port Ethernet switch associated with it that lets traffic either pass
through, or to be injected/removed by the device.
06/15/15
CableLabs®
65
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
10.2.1 Backhaul Daisy-chaining Requirements
The requirements regarding backhaul daisy-chaining are:
•
Each RPD that is to be individually authenticated MUST have its own MAC address and its own IP address
assignment.
•
When operating in a daisy-chained topology, the RPD MUST support the authentication requirements defined in
Section 6.3.
10.3 Networking Considerations
It is important to distinguish between the terms “PHB-ID” and “DSCP” as used in the MHAv2 specifications:
•
A “PHB-ID” is a 6-bit value appearing in an L2TPv3 Attribute Value Pair (AVP)
•
A “DSCP” is a 6-bit value appearing in an IP packet header
All L2TPv3 packets in [R-DEPI] are in a control session, a PSP session, or a non-PSP session. All PSP sessions
contain both downstream and upstream data “flows”. PHB-IDs apply to flows of PSP data sessions. DSCPs apply to
the IP packets that encapsulate all L2TPv3 packets, i.e., DSCPs apply to control sessions, PSP sessions, and non-PSP
sessions.
For a downstream PSP flow, the CCAP-Core assigns via L2TPv3 AVPs the PHB-ID for each downstream PSP flow.
The assigned downstream flow PHB-ID selects the scheduling behavior for that flow only on the RPD, i.e., for the
scheduling of multiple downstream PSP flows on the single hop from CIN to RF network. At a minimum, an RPD
SHOULD provide highest-priority strict priority service to PSP flows assigned to the Expedited Forwarding(46)
PHB-ID. Support for more complex scheduling disciplines, e.g. multiple strict priorities or weighted fair queueing, is
for further study.
The RPD advertises via GCP to the CCAP-Core what PHB-IDs it supports for downstream PSP flow scheduling. An
RPD MUST support at least the Expedited Forwarding (46) and BestEffort(0) PHB-IDs. The use of PHB-IDs other
than ExpeditedForwarding(46) and BestEffort(0) is for further study.
For a downstream PSP flow, The CCAP-Core selects the DSCP to send in the IP header of the L2TPv3 data session
packets for the flow. The DSCP selects the per-hop behavior on each CIN router between the CCAP-Core and RPD.
The 6-bit DSCP of the IP headers of downstream L2TPv3 data packets on a PSP flow may or may not equal the 6-bit
PHB-ID assigned to the flow on the RPD itself. For example, the CCAP-Core may use more than two different DSCP
values when the CIN supports them. The RPD ignores the DSCP of a downstream IP packet and uses only the flow ID
in the inner PSP sub-layer to select the queue with which it schedules downstream data for the flow.
For an upstream PSP flow, the CCAP-Core assigns via L2TPv3 AVPs the PHB-ID for each upstream PSP flow. For
the upstream case, the PHB-ID corresponds to a “recommended DSCP value” as described in [RFC 3140]. The RPD
sets the DSCP in the IP headers of all upstream L2TPv3 data packets for a PSP flow to the PHB-ID value assigned to
that flow. The PHB-ID assigned to an upstream PSP flow does not identify any per-hop beviour in the RPD itself.
10.3.1
Per Hop Behavior
The IETF has defined a number of Per Hop Behaviors (PHBs) to be used for offering network-based QoS. DEPI
supports use of the 6-bit Expedited Forwarding (EF) PHB as described in [RFC 3246], Assured Forwarding (AF)
PHBs as described in [RFC 2597], and best effort forwarding as described in [RFC 2597]. DEPI negotiates six-bit
Per-Hop Behavior Identifiers (PHBIDs) between the CCAP-Core and the RPD.
The RPD advertises the PHB-IDs it supports for its downstream PSP packet scheduler. The RPD MUST support
Expedited Forwarding(46) and BestEffort(0) PHB-IDs. The RPD SHOULD provide highest strict priority scheduling
service to PSP flows assigned to the Expedited Forwarding(46) PHB-ID. The CCAP-Core SHOULD support
assigning the Expedited Forwarding(46) PHB-ID to a separate PSP flow for MAPs+UCDs with Best Effort (0) for all
other traffic.
For upstream flows, the RPD MUST support signaling of an arbitrary 6-bit PHB-ID as the transmitted 6-bit DSCP
value.
66
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
NOTE: Table 1 lists the PHBs explicitly supported by the DEPI specification. This specification does not prohibit
support for other PHBs not defined in Table 1.
Table 1 - PHBs and Recommended DSCP Values
PHB
PHB ID(s) and Recommended DSCP Value(s)
EF
46
AF (multiple levels)
10, 12, 14, 18, 20, 22, 26, 28, 30, 34, 36, 38
Best effort
0
The DEPI interface supports multiple traffic types including DOCSIS MAC and DOCSIS data traffic. Within both
traffic types, there may be different levels of priority. For PSP operation, the CCAP-Core SHOULD provide a
mechanism to map traffic of different priorities to DEPI flows with different PHB values. The CCAP-Core SHOULD
NOT use the same PHB across multiple DEPI flows within a session.
The CIN should provide the appropriate Per Hop Behavior for the differentiated traffic types. The level of granularity
provided for differentiated traffic is determined by the network operator, but at a minimum, it is expected that
DOCSIS MAP messages and VoIP data traffic are prioritized higher than best effort data traffic.
The RPD uses the PHB signaled in the establishment of the DEPI flow when scheduling multiple DEPI PSP flows
onto one QAM channel as described in [R-DEPI].
10.3.2
DiffServ Code Point Usage
An operator sets up CIN network elements to support a particular set of DSCPs. The selected DSCPs should select
appropriate per-hop behavior at each network element for differentiated traffic types.
For L2TPv3 data sessions, packets in the same direction have the same DSCP; packets in different directions may have
different DSCPs. For PSP L2TPv3 sessions, each PSP “flow” in the session and in a particular direction may have a
different DSCP value. Different PSP flows in the same PSP session may have the same DSCP.
The CCAP-Core is responsible for selecting the DSCP values of all L2TPv3 control and data session packets,
including the DSCP sent by the RPD.
The CCAP-Core and RPD MUST:
•
set the same DSCP on all L2TPv3 control packets;
•
set the same DSCP on all L2TPv3 data packets of the same non-PSP session in a direction;
•
set the same DSCP on all L2TPv3 data packets of the same PSP flow in a direction.
DOCSIS frames encapsulated in L2TPv3 packets may contain IP packets which also have a DSCP assigned. The RPD
is not required to schedule packets based upon the original DSCP contained within the DOCSIS frame.
10.3.3
Packet Sequencing
For a stream of packets transmitted on a DEPI flow, the packet sequence number is incremented by one for each packet
sent, as described in [R-DEPI].
If the RPD detects a discontinuity in the packet sequence numbers indicating that one or more packets were dropped or
delayed, an error is logged and the RPD SHOULD transfer the current packet to the QAM channel without waiting for
the missing packets. If the RPD detects a discontinuity in the packet sequence numbers indicating that one or more
packets have arrived late, those packets SHOULD be discarded.
The RPD MUST NOT forward packets that were skipped due to a discontinuity in the sequence numbers. Storing and
re-ordering of packets so that they can be delivered to the QAM channel in the correct sequence is not prohibited by
these requirements and the RPD MAY perform such re-ordering as long as the latency requirements of Section 5.6 are
met.
06/15/15
CableLabs®
67
CM-SP-R-PHY-I01-150615
10.3.4
Data-Over-Cable Service Interface Specifications
Network MTU
The network between the CCAP-Core and the RPD has a certain Maximum Transmission Unit (MTU). If a maximum
size DOCSIS frame were to be tunneled from the CCAP-Core to the RPD without fragmentation, the size of the
resulting packet could be greater than the CIN can handle. Both the D-MPT and PSP modes avoid this issue by
offering streaming and fragmentation. As such, IP fragmentation is not required. IP fragmentation is also undesirable
because the RPD may forward packets based upon the destination UDP port, and the UDP port is only available in the
first IP fragment.
Determining the MTU to use for the L2TPv3 tunnel between the CCAP-Core and the RPD is a two-step process:
1.
Choose the payload size.
2.
Determine the MTU of the path between the CCAP-Core and the RPD.
The first step is done as part of L2TPv3 session establishment (see [R-DEPI]) using the DEPI MTU AVPs. When the
CCAP-Core sends the session ICRQ message it MUST supply the DEPI Local MTU AVP with a payload size that is
the lesser of its receive capabilities and the receive capabilities defined by its lower layer. The receive capabilities of
the CCAP-Core are defined by its internal constraints, and any configured maximums. The receive capabilities
defined by its lower layer are calculated based on referencing the payload size constraints of the interface below which
this tunnel is being created, as defined in Annex A.1.
The CCAP-Core MUST support an MTU size of at least 2000 bytes, as calculated in Annex A.1. The RPD MUST
send L2TPv3 frames with a payload size less than or equal to this maximum. If the RPD cannot meet this criterion then
it MUST fail session creation by generating a CDN message. The RPD needs to consider the same criteria in
calculating its MTU.
The RPD MUST support an MTU size of at least 2000 bytes, as calculated in Annex A.1. The RPD MUST insert the
DEPI Remote MTU AVP in the ICRP message with its MTU size. The CCAP-Core MUST send L2TPv3 frames with
a payload size less than or equal to this maximum. If the CCAP-Core cannot meet this criterion then it MUST fail
session creation by generating a CDN message.
The second step is to determine the MTU of the path between the CCAP-Core and the RPD. The CCAP-Core MUST
provide a mechanism to prevent sending packets larger than the network MTU. This SHOULD be done using Path
MTU Discovery, as described in [RFC 1191]. Annex A.3 gives a brief overview of the Path MTU discovery protocol.
Alternatively, this MAY be done via a static configuration option. Both the CCAP-Core and the RPD MUST have a
way to statically configure an MTU for each L2TPv3 session. To avoid IP fragmentation, the CCAP-Core and the
RPD MUST set the Don't Fragment bit (DF) in the IPv4 header for all transmissions into the L2TPv3 pseudowire.
10.4 Pilot Tone Generation
Pilot tones in the HFC network are required to ensure that amplifiers in the network are operating correctly. Amplifiers
use these tones to adjust gain and keep signals at the appropriate output level. The RPD in a distributed CCAP
architecture is required to be capable of generating these tones and placing these tones in the appropriate portion of the
downstream spectrum under control of the CCAP-Core. Pilot tones, in this context, are used for Automatic Gain
Control (AGC) and should not be confused with the pilots used in OFDM/OFDMA channels.
The RPD MUST support pilot tone generation.
The RPD MUST support up to four frequencies for CW pilot tones per chassis simultaneously.
The RPD MUST support a minimum of two CW pilot tones output per downstream RF port.
The RPD MUST support placing CW pilot tones from 54 to 535 MHz.
The RPD SHOULD support placing CW pilot tones from 54 to 1218 MHz.
The RPD MUST allow the pilot tone to be set in a range of +3 dB to +9 dB in 0.2 dB steps relative to the 256-QAM
level.
The RPD SHOULD allow pilot tones to be set in absolute dBmV across that range.
68
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
The RPD MUST support a pilot tone power accuracy of ±2 dB. Time and temperature stability requirements are
expected to be defined in a product specification document.
The RPD MUST support a minimum pilot tone quality (e.g., 2nd and 3rd harmonics, 2nd and 3rd inter-modulation
distortion, spurs, etc.) of 65 dBc relative to the pilot tone output power, in the case where Neq’ ≥ Neq/4 where:
•
Neq’ = Number of equivalent active 6 MHz channels combined per RF port
•
Neq = Number of equivalent channels the RPD is capable of per RF port
The RPD MUST meet the requirements for noise in other channels (47 MHz to 1002 MHz) as defined in the CMTS
Output Out-of-Band Noise and Spurious Emissions Requirements table of [PHY v3.1] for the given Neq and Neq’, with
the exception of channels coinciding with the pilot tones’ harmonics products. For this purpose, channels occupied by
pilots are considered active channels, but the noise is calculated relative to the 6 MHz equivalent channels’ power.
06/15/15
CableLabs®
69
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
11 MULTIPLE CCAP-CORE OPERATION
11.1 Introduction
The MHAv2 architecture permits RPDs to be managed by more than one CCAP-Core. An RPD is controlled by
exactly one “principal” CCAP-Core and zero or more “auxiliary” CCAP-Core(s). An “auxiliary” core manages a
subset of RPD resources, e.g., particular channels or RF ports. Each auxiliary CCAP-Core establishes its own GCP
session and L2TPv3 control sessions with the RPD. The specification term “CCAP-Core” can refer to either the
principal core or an auxiliary core.
Potential auxiliary CCAP-Cores include but are not limited to the following:
•
A “Broadcast EQAM” CCAP-Core that controls only downstream video broadcast channels;
•
A “Narrowcast EQAM” CCAP-Core that controls only downstream video narrowcast channels;
•
A “Forward OOB” CCAP-Core that controls and sources NDF, typically broadcast to multiple RPD ports;
•
A “Reverse OOB” CCAP-Core that controls and receives NDR channels, always received one per RPD port.
•
A CMTS CCAP that controls the downstream and upstream channels of a separate MAC domain;
A CMTS Core is programmed or configured in a vendor-specific manner to operate as an auxiliary core.
11.2 RPD Startup with Multiple Cores
An RPD MUST implement an ActivePrincipalCore object into which the principal CCAP-Core controlling the RPD
writes its IP address.
Attribute Name
ActivePrincipalCore
Type
IpAddress
Access
Type Constraints
Units
Default
RW
An RPD MUST implement an ActiveAuxCoreTable object into which auxiliary CCAP-Cores connected to the RPD
write their IP addresses.
Attribute Name
Type
ActiveAuxCoreTable
ActiveAuxCoreIp
Access
Type Constraints
Units
Default
RW
IpAddress
RW
key
An RPD MUST implement a ConfiguredCoreTable object that contains the list of principal and auxiliary cores to
which the RPD attempts to attach. This table is originally populated by the RPD itself based on DHCP, but may be
modified by the principal core.
Attribute Name
Type
ConfiguredCoreTable
ConfiguredCoreIp
Access
Type Constraints
Units
Default
RW
IpAddress
RW
key
A resetting RPD MUST clear its ConfiguredCoreTable, ActivePrincipalCore and ActiveAuxiliaryCoreTable objects.
This requirement applies to both cold-reset and warm-reset.
An RPD MUST accept up to six CCAP-Core-IP-Address options (a new [CANN] 42.x option) in its DHCP response.
A starting RPD initially populates its ConfiguredCoreTable with the CCAP-Core-IP-Address list learned from DHCP.
After completing initial DHCP, an RPD MUST attempt to establish EAP-TLS authentication and a GCP TCP session
with each of the ConfiguredCoreTable IP addresses until a Principal core identifies itself by writing to the
ActivePrincipalCore RPD object.
After a GCP TCP session is established, the Principal CCAP-Core MUST write its IP address into the
ActivePrincipalCore object of the RPD before attempting any other GCP write operations.
70
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
After a GCP TCP session is established, an auxiliary CCAP-Core MUST add its IP address to the RPD’s
ActiveAuxiliaryCores table. The Principal core MAY add IP addresses to the ActiveAuxiliaryCores table.
The RPD MUST attempt to maintain an IPsec-authorized GCP session to each core IP address in its
ActivePrincipalCore object and ActiveAuxiliaryCores table.
11.3 Downstream Channel Constraint Table
Downstream QAM modulators are often implemented with hardware “channel blocks” that constrain consecutively
identified channels to have the same or related physical attribute. A Principal CCAP-Core attempting to dynamically
determine resource sets of downstream channels is made aware of those constraints by a read-only object table on the
RPD.
An RPD MUST implement a read-only DownChannelConstraintTable to identify constraints imposed by its hardware
on the configuration of physical parameters of blocks of downstream channels.
The set of described constraints are implied as present on all downstream RF ports of the RPD.
Attribute Name
DownChannelConstraintTable
Type
Access
Type Constraints
Units
Default
Na
Index
unsignedInt
DownChanIndexStart
unsignedInt
key
0..159
DownChanIndexEnd
unsignedInt
0..159, >
DownChanIndexStart
LockParameters
LockParamBits
The constraints are defined as a LockParameters bitmask:
LockParameters EnumBits {
frequency(0),
bandwidth(1),
power(2),
modulation(3),
interleaver(4),
j83Annex(5),
symbolRate(6)
mute(7)
}
The LockParameters field is a bitmask from which constraints apply to the range of channels defined by
DownChanIndexStart through DownChanIndexEnd, inclusive. Note that different DownChannelConstraintTable
objects may describe different LockParameter values on overlapping or partially overlapping channel ranges of other
DownChannelContraintTable objects:
“frequency(0)” means the channels are constrained to have consecutive frequencies;
“bandwidth(1)” means the channels are constrained to have the same channel width;
“power(2)” means the channels are constrained to have the same power adjustment;
“modulation(3)” means the channels are constrained to have the same modulation;
“interleaver(4)” means the channels are constrained to have the same interleave value;
“j83Annex(5)” means the channels are constrained to have the same j83Annex definition;
“symbolRate(6)” means the channels are constrained to have the same symbol rate;
“mute(7)” means the channels are constrained to be muted or unmuted together.
06/15/15
CableLabs®
71
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
11.4 Resource Sets and Auxiliary Resource Assignment
An RPD MUST implement the ResourceSet table , which identifies which auxiliary cores may control which RPD
object.
Attribute Name
Type
Access
Type Constraints
Units
Default
ResourceSet
ResourceSetIndex
Integer
RW
key
DsRfPortStart
Integer
RW
-1 if unused
DsRfPortEnd
Integer
RW
-1 if unused
DsChanIndexStart
Integer
RW
-1 if unused
DsChanIndexEnd
Integer
RW
-1 if unused
UsRfPortStart
Integer
RW
-1 if unused
UsRfPortEnd
Integer
RW
-1 if unused
UsChanIndexStart
Integer
RW
-1 if unused
UsChanIndexEnd
Integer
RW
-1 if unused
A resource set consists of a range of channels from start to end (inclusive) on particular RF ports from start to end
(inclusive). The RPD MUST enforce that no two entries in the ResourceSet table overlap, i.e., include the same
channel on the same RF port.
A Principal CCAP-Core MUST implement R-OSSI tables that permit CCAP-Core configuration of the ResourceSet
table objects to be written to an RPD.
An RPD MUST implement the AuxResourceAssignment table, which identifies which auxiliary core is authorized to
manage which resource set. The RPD MUST implement sufficient entries in AuxResourceAssignment to assign each
supported auxiliary core to one resource set.
Attribute Name
AuxResourceAssignment
Type
Access
Type Constraints
Units
Default
List
ResourceSetIndex
key
RW
AuxCoreIpAddress
key
RW
See text.
The RPD MUST permit more than one AuxCoreIpAddress to be configured to the same ResourceSetIndex.
An RPD MUST implement the PermitAuxSelfConfiguration object to control whether auxiliary cores are
permitted to configure their own resource sets.
Attribute Name
PermitAuxSelfConfiguration
Type
Boolean
Access
RW
Type Constraints
Writeable by Principal
core only
Units
Default
false
An RPD MUST enforce the policy that only the Principal core can write to PermitAuxSelfConfiguration.
The Principal core is by default solely responsible for writing the ResourceSet and AuxResourceAssignment tables.
When PermitAuxSelfConfiguration is ‘false’ (the default), the RPD MUST enforce that only the Principal core
may write to the ResourceSetTable and AuxResourceAssignmentTable. When PermitAuxSelfConfiguration
is ‘true’ (as changed by the Principal core) the RPD MUST permit any auxiliary core to write to ResourceSetTable
and to assign those entries to its own IP address by writing to the AuxResourceAssignment table. Even when
PermitAuxSelfConfiguration is ‘true’, the RPD MUST enforce the policy that an auxiliary core writes only its
own IP address into the AuxResourceAssignment table.
72
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
11.5 RPD Reads and Writes
The RPD MUST support concurrent reads of any objects by any connected CCAP-Core, whether a Principal core or an
Auxiliary core.
The RPD MUST reject an attempt by an Auxiliary core to write to any object not specifically authorized to it by the
AuxResourceAssignment table. The RPD MUST reject an attempt by the Principal core to write to any object
assigned to an Auxiliary core in the AuxResourceAssignment table.
06/15/15
CableLabs®
73
CM-SP-R-PHY-I01-150615
Annex A
A.1
Data-Over-Cable Service Interface Specifications
DEPI MTU (Normative)
L2TPv3 Lower Layer Payload Size
Typically, an interface calculates its default maximum payload size by asking the interface below it in the interface
channel what is its maximum payload size and considering its own encapsulation. For example, by default, Ethernet
has a frame size of 1518 (without VLANs). The Ethernet encapsulation is 18 bytes, leaving 1500 bytes of payload
(MTU) for its upper layer. IP then subtracts the IP header size (typically 20 bytes) to arrive at 1480 bytes available to
its upper layer. For D-MPT the remainder becomes 1472 bytes, because the Session Field and the L2TPv3 Data
Session Header comprise 8 bytes. For PSP, the PSP header including the maximum PSP segment table size needs to be
taken into account.
The CCAP-Core and the RPD MUST support expanded Ethernet Frame sizes, up to 2000 bytes long, in compliance
with [MULPI v3.1].
A.2
Maximum Frame Size for DEPI
This section documents the maximum frame size of the DEPI when a PSP pseudowire is used without fragmenting or
concatenation.
Table 2 - MTU of DEPI (for PSP)
Field
Size
Ethernet Header
14 bytes
4 bytes
20 bytes
IPv6 Header
40 bytes
L2TPv3 Header
8 bytes
DEPI-PSP Header***
6 bytes
DOCSIS Frame
IPv4 Header
DEPI MTU
DEPI Frame
802.1Q Header
DOCSIS Header****
6-246 bytes
Ethernet Header
14 bytes
802.1Q Header
4 bytes
Ethernet PDU
1500 or 2000 bytes
Ethernet CRC
4 bytes
Ethernet CRC
4 bytes
Total with PSP, no UDP, IPv4, no VLAN
1570 to 1862 (or 2362)
*** A PSP header is 4 bytes plus 2 bytes for each segment. Only one segment is shown. (A D-MPT
header is 4 bytes.)
**** A typical DOCSIS header with BPI and no other extended headers is 11 bytes.
For simplicity, only one PSP segment is included in the above calculations. Additional segments are needed when PSP
is concatenating or fragmenting. Note that a 2000 byte payload in a PSP frame could contain as many as 26
uncompressed TCP ACKs (64 byte Ethernet packets plus 6 to 11 bytes of DOCSIS overhead) which could create as
many as 22 segments (first and last packets are fragmented) which would create a segment table size of 44 bytes, in
addition to the standard 4 byte PSP header. For other payload types such as VoIP packets with high codec compression
and with PHS disabled, or with larger MTUs, the number of segments could be even higher.
A.3
Path MTU Discovery
Path MTU Discovery relies on the fact that the network elements between the CCAP-Core and the RPD all support
this functionality [RFC 1191]. If these network elements do not support Path MTU Discovery then this mechanism
cannot be used and the static configuration option should be used.
Path MTU Discovery (PMTUD) works when the IP path MTU between the CCAP-Core and the RPD is less than the
total IP datagram size generated when using the payload size negotiated during L2TPv3 session establishment, and the
74
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
Don’t Fragment (DF) bit is set in the IP header. If the CCAP-Core sends packets larger than the network can support,
then network elements between the CCAP-Core and the RPD may generate an ICMP Destination Unreachable
message with the code "Fragmentation needed and DF set" (ICMP Type 3 Code 4, also referred to as "Datagram Too
Big" message), toward the source of the tunneled packet, if ICMP unreachables are allowed.
This ICMP error message includes at least the IP header and the next 8 bytes of the IP data (corresponding to the UDP
header when using L2TPv3 over UDP, or to the Session ID and first 4 bytes of the L2SS when using L2TPv3 over IP)
from the offending packet. The CCAP-Core and the RPD should have a way to map the source and destination IP
address contained in the IP header embedded in the ICMP data to an L2TP Control Connection. As defined in [RFC
1191], a "PMTU is associated with a path, which is a particular combination of IP source and destination address and
perhaps a Type-of-Service (TOS)".
Upon successfully processing the ICMP Destination Unreachable message, the CCAP-Core and RPD should reduce
the Max Payload of all the sessions associated with the control connection mapped from the ICMP Destination
Unreachable message to the size requested in the Next-Hop MTU field of the message. Both the Max Payload and the
size contained in the Next-Hop MTU field express a Layer 3 payload of a Layer 2 frame, including the IP header and
IP data.
The Max Payload MUST NOT be increased by receiving an ICMP Destination Unreachable message. The
CCAP-Core and RPD may periodically attempt to increase the Max Payload of the session to its negotiated maximum
and restart this process in case the path through the network has changed and larger MTUs are allowed. This technique
is described in [RFC 1191]. The Max Payload size learned through this process will never be greater than the
negotiated maximum learned during session establishment. The Path MTU Discovery procedures for IPv6 are
described in [RFC 1981].
06/15/15
CableLabs®
75
CM-SP-R-PHY-I01-150615
Annex B
Data-Over-Cable Service Interface Specifications
GCP Usage (Normative)
GCP (Generic Control Plane) is described in [GCP]. GCP is fundamentally a control plane tunnel that allows data
structures from other protocols to be reused in a new context. This is useful if there is configuration information that is
well defined in an external specification. GCP can repurpose the information from other specifications rather than
redefining it. For example, MHAv2 uses GCP to reuse predefined DOCSIS TLVs for configuration and operation of
the RPD. GCP has three basic features:
•
Device management, such as power management;
•
Structured access, such as TLV tunneling;
•
Diagnostic access.
GCP defines the structured access using a combination of:
•
32 bit Vendor ID as defined in [Vendor ID];
•
16 bit Structure ID as uniquely defined by the vendor. For MHAv2, the default vendor ID is the CableLabs vendor
ID of 4491 (decimal).
When GCP tunnels the data structures of another protocol, the syntax GCP(protocol name) can be used. For example,
if the DOCSIS UCD command is tunneled over GCP, the combination of the DOCSIS UCD over GCP can be referred
to as GCP(UCD).
B.1
RPD Configuration with GCP(UCD)
To configure the RPD upstream PHY, the DOCSIS UCD command is tunneled through GCP using the following GCP
parameters:
•
Vendor ID = 4491 (CableLabs)
•
Structure ID = 35 (DOCSIS UCD)
The UCD is sent unicast to an RPD with a GCP Exchange Data Structures Request message. The RPD responds with
a GCP Exchange Data Structures Response message.
B.2
RPD Upstream Scheduler with GCP(DSx)
MHAv2 permits the upstream scheduler to be located either centrally in the CMTS Core or in the RPD. When the
scheduler is located in the RPD, the CMTS Core needs to be able to add, change, and delete service flows in the remote
upstream scheduler. The semantics for doing this are fully described in the DOCSIS DSA (Dynamic Service Flow
Add), DSC (Dynamic Service Flow Change), and DSD (Dynamic Service Flow Delete) commands.
These commands are tunneled through GCP with the following parameters:
•
Vendor ID = 4491 (CableLabs)
•
Structure ID as defined in Table 3.
The DSx command headers are not needed, because GCP contains all header information and a transaction ID. The
specific payload of the DSx commands that are used in the corresponding GCP commands are shown in Table 3. GCP
does not have a separate ACK command since GCP is transported over a reliable transport protocol such as TCP. If the
DSx-ACK TLVs are needed, they are carried over a second GCP Request Response pair.
Table 3 - GCP Encoding for the Upstream Scheduler
76
Structure ID
Function
GCP Message
GCP Payload
15
DSA-REQ
EDS-REQ
TLVs
16
DSA-RSP
EDS-RSP
Confirmation Code, TLVs
17
DSA-ACK
EDS-REQ
Confirmation Code, TLVs
17
n/a
EDS-RSP
No content
18
DSC-REQ
EDS-REQ
TLVs
CableLabs®
06/15/15
Remote PHY Specification
B.3
CM-SP-R-PHY-I01-150615
Structure ID
Function
GCP Message
GCP Payload
19
DSC-RSP
EDS-RSP
Confirmation Code, TLVs
20
DSC-ACK
EDS-REQ
Confirmation Code, TLVs
20
n/a
EDS-RSP
No content
21
DSD-REQ
EDS-REQ
SFID, TLVs
22
DSD-RSP
EDS-RSP
Confirmation Code
R-PHY Control Protocol
The following section defines the rules for the application of GCP as a Remote PHY control plane protocol. This set of
rules is referred to as R-PHY Control Protocol or RCP.
RCP operates as an abstraction layer over the foundation of GCP protocol as defined in [GCP]. RCP provides the set
of CCAP-Core with to ability to remotely manage a set of objects, such as channels, ports, performance variables, etc.
RCP relies on the following GCP messages: Notify, Device Management and Exchange Data Structures. The
encodings of the GCP messages are provided in tables below.
B.3.1
RCP over GCP EDS Message
Table 4 shows the encodings of the RCP over GCP EDS message.
Table 4 - RCP Encodings for GCP EDS Messages
Description
Message ID
1 byte
Contents
6 (Exchange Data Structures Request)
Message Length 2 bytes
12 + N (length excludes three first bytes (Id +Len) of the header)
Transaction ID
2 bytes
Unique value
Mode
1 byte
0
Port
2 bytes
N/A
Channel
2 bytes
N/A
Vendor ID
4 bytes
4491 (IANA Enterprise Number assigned to CableLabs)
Vendor Index
Message Body
B.3.2
Length
1 byte
N bytes
1
TLV encoded RCP Message
RCP over GCP EDS Response Messages
The EDS Normal Response message shown in Table 5 has a format identical to the Request message (except Message
ID == 7) and permits the inclusion of the TLV-encoded information.
Table 5 - RCP Encodings for GCP EDS Normal Response Messages
Description
Message ID
06/15/15
Length
1 byte
Contents
7 (Exchange Data Structures Request Normal Response)
Message Length 2 bytes
12 + N (length excludes three first bytes (Id +Len) of the header)
Transaction ID
2 bytes
Unique value, same as request
Mode
1 byte
0
Port
2 bytes
N/A
Channel
2 bytes
N/A
Vendor ID
4 bytes
4491 (IANA Enterprise Number assigned to CableLabs)
Vendor Index
1 byte
1
CableLabs®
77
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
Description
Length
Message Body
N bytes
Contents
TLV encoded RCP Message
The EDS Error Response (Message Id == 8) format shown in Table 6 does not include TLV encoding information.
This message can be used to communicate errors in those cases which are defined by the GCP specification [GCP].
The types of errors which are not covered by GCP Error Response Message are conveyed in EDS Normal Response
Message in TLV-encoded format.
Table 6 - RCP Encodings for GCP EDS Error Response Messages
Description
Message ID
B.3.3
Length
Contents
1 byte
135 (Exchange Data Structures Error Response)
Message Length 2 bytes
3
Transaction ID
2 bytes
Same as request
Exception code
1 byte
See section 6.4 of [GCP]
RCP over GCP Device Management Message
The RCP encodings of GCP Device Management messages are shown in Table 7.
Table 7 - RCP Encodings for GCP Device Management Messages
Description
Message ID
Length
1 byte
Message Length 2 bytes
Contents
4 (Device Management)
8
Transaction ID
2 bytes
Unique value
Mode
1 byte
Bit 7: 0 = Send normal response
1 = Suppress normal response
Bit 6-0: Reserved. Set to 0.
Port
2 bytes
N/A
Channel
2 bytes
N/A
Command
N bytes
0 – Null
1 – Cold Reset
2 – Warm Reset
3 – Standby
4 – Wakeup
5 – Power-Down
6 – Power-Up
7 to 255 – Reserved
The RPD MUST set bit 7 of the Mode field to ‘1’.
78
CableLabs®
06/15/15
Remote PHY Specification
B.3.4
CM-SP-R-PHY-I01-150615
RCP over GCP Notify Message
GCP Notify messages are sent from the RPD to the CCAP-Core. RCP utilizes Event Code 1 and the TLV-encoded
portion of the GCP Notify message. CCAP-Core does not respond to Notify messages.
The RPD MUST set bit 7 to ‘1’ and bit 6 to ‘1’ in the Mode field. The RPD MUST set the value of the Event Code field
to “1”.
The RCP encodings of GCP Notify messages are shown in Table 8.
Table 8 - RCP Encodings for GCP Notify Messages
Description
Message ID
B.3.5
Length
1 byte
Contents
2 (Notify)
Message Length 2 bytes
8 + N (length does not include first 3 bytes of the message)
Transaction ID
2 bytes
Unique value, selected by the RPD.
Mode
1 byte
Bit 7: 0 = Send normal response
1 = Suppress normal response
Bit 6: 0 = Event data is text
1 = Event data is raw
Bit 5-0: Reserved. Set to 0.
Status
1 byte
0 – Null (default)
1 – Cold Reset
2 – Warm Reset
3 – Standby
4 – Wakeup
5 – Power-Down
6 – Power-Up
7 to 255 – Reserved
Event Code
4 bytes
1
Event Data
N bytes
TLV-encoded RCP message
RCP TLV Format
The information carried in RCP protocol is formatted into TLV tuples. RCP operates with TLV format and usage rules
which are similar to those defined in DOCSIS protocol. Each RPC TLV consists of a one byte long Type field, two
byte long Length field and an optional, variable length Value field. The RPC TLV Type field can have the value of
1-255. The use of the value of “0” is reserved. The RPC TLV Length field denotes the total length of the Value field.
The valid range for the Length field is 0-65535. When a TLV does not include the Value field, the Length field is set to
zero. The encoding of the Value field varies, depending on the type field. The RCP protocol allows for nesting of
sub-TLVs.
Using these encodings, new parameters may be added which some devices cannot interpret. A CCAP-Core or RPD
which does not recognize a parameter type MUST skip over this parameter and not treat the event as an error
condition.
B.3.6
RCP Message Structure
The RCP Messages are embedded in a single TLV tuple. The value field of these TLV consists of multiple tuples in the
form {operation-TLV, Object Set-TLV}. The RCP protocol defines three operation types: “Read”, “Write”, and
“Delete”. The definition of the managed objects, also referred to as information model or schema is provided further in
this specification.
The RCP TLV format imposes a size limit on RCP messages of 64 kB. RCP messages are never fragmented. When
necessary, for example if the volume of information exceeds 64 kB, the CCAP-Core can issue multiple messages.
06/15/15
CableLabs®
79
CM-SP-R-PHY-I01-150615
B.3.7
Data-Over-Cable Service Interface Specifications
RCP Messages Types
The RCP protocol defines three message types. These messages, their TLV encoding, description and GCP usage are
presented in Table 9.
Table 9 - Summary of RCP Messages
Message Name
Message
TLV Type
Description
GCP Mapping
IRA, Identification and
Resource Advertising
01
An initial message exchanged after authentication in which
the CCAP-Core obtains all parameters identifying the RPD
and its available resources.
Sent by CCAP-Core in GCP
EDS message.
REX, RCP Object
Exchange
02
A message in which CCAP-Core allocates or de-allocates
resources and configures the resources in the RPD or
requests information from the RPD i.e. statistics or other
status data.
Sent by the CCAP-Core in
GCP EDS message.
Responded to by the RPD
when operation is complete.
NTF, Notification
03
A message sent by the RPD to inform the CC about a specific
event or a set of events.
Sent by the RPD in GCP
Notify Message. CC does
not respond to NTF
messages.
B.3.8
RCP Protocol Rules
The CCAP-Core can issue multiple RCP messages before it receives acknowledgement from the RPD. Each RPD
MUST support a minimum of 16 outstanding messages per CCAP-Core. A CCAP-Core MAY issue a single IRA or
REX message with a combination of read, write and delete tuples. The NTF issue by the RPD may only contain write
tuples. A CCAP-Core may issue a “Read” operation for a set individual objects (leaves) or object trees.
Responses to IRA and REX messages indicate the result of request processing with granularity of each
{operation-TLV, Object Set-TLV} tuple. When the RPD response indicates a failure for a particular tuple, the RPD
MUST make no change to the objects indicated in the tuple.
The RPD MUST respond to RCP request messages with one second of receiving the request message. The response
messages sent by the RPD may be issued in a different order from the order of reception of request messages.
Since GCP operates over a reliable TCP connection) the protocol does not define explicit “acknowledgement”
messages or other mechanism to deal with loss of individual messages.
B.3.9
Extensibility
This section will be written for a future version of this specification.
B.3.10 Protocol Versioning
The RCP protocol uses versioning as the primary means for future extensibility. The initial RCP protocol version
defined by this specification is “1.0”. Future versions of this specification may define new RCP protocol versions with
additional capabilities or protocol options. During the initialization the CCAP-Core will read the RPD’s capabilities,
including the set of RCP protocol versions supported by the RCP via the IRA message. The CCAP will then select the
highest RCP protocol version that both the CCAP-Core and the RPD can support and instruct the RPD to use the
selected version.
B.3.11 Information Model Extensibility
The RPHY information model/schema is versioned separately from the protocol. The method for schema version
selection is similar to the protocol version selection. The initial RCP information schema version defined by this
specification is “1.0” Future versions of this specification may define new RCP information schema versions. For each
version of the schema this specification will define a set of mandatory objects and a set of optional objects organized in
sets, referred to as features. During initialization, the CCAP-Core will read which schema features the RPD supports
in the IRA message. The CCAP-Core will also let the RPD know (write) which versions of the schema and which
features it supports to control objects sent in Notify messages.
80
CableLabs®
06/15/15
Remote PHY Specification
CM-SP-R-PHY-I01-150615
The CCAP-Core MUST convey in RCP protocol only those objects that the RPD supports. RPD MUST convey in
RCP protocol only those objects that the CCAP-Core supports. These requirements are not applicable to vendor
specific extensions.
B.3.12 Vendor Specific Extensions
The RCP protocol permits for exchange of vendor specific information by defining a method for inclusion of vendor
specific TLVs. Vendor specific TLVs are complex TLVs with a Type of “Vendor-Specific”. The first sub-TLV of a
vendor specific TLV is the TLV identifying the vendor with length of 4 and the value field containing the vendor’s
Private Enterprise Number (http://www.iana.org/assignments/enterprise-numbers/enterprise-numbers). A vendor
specific TLV includes one or more vendor defined sub-TLVs. The definition of the formats and the usage of these
TLVs are outside of the scope of this specification.
An example of vendor specific TLV is provided below.
{T= Vendor-Specific, Length: variable (minimum)
{T = Vendor-Id, L = 4, V = Vendor ID: Enterprise number identifying vendor}
{
A sequence consisting of one or more vendor specific TLVs
}
}
Vendor- specific TLVs are ignored by RPDs and CCAP-Cores which do not recognize vendor id.
B.3.13 Inclusion of DOCSIS Messages
The CCAP-Core can include in RCP certain messages describing the majority of the parameters of US TDMA and
OFDMA channels and DS OFDM channels. These messages are transmitted in the form of an TLVs in REX
messages.
The RPD MUST support the reception of three types of DOCSIS messages, including UCD, OCD and DPD Messages
as the means for configuration of the DOCSIS channels for which these messages provide description. The RPD
MUST decode these messages using DOCSIS rules in order to configure certain channel resources.
An example of an RCP message containing an embedded DOCSIS message is provided in Section B.3.14.4.
B.3.14 RCP Message Examples
B.3.14.1
RCP Rex Message Request Example
The following example presented below represents a message with a single “Read” operation for a set of statistical
counters for two upstream channels. Curly braces “{“ and “}” denote the boundaries of TLVs. Note, that the outer
envelope (GCP EDC Request) is not shown.
{ T = REX, L= 43, V=
; top-level “container” type
{ T = Sequence, L = 40, V =
; a seq. of TLVs starting with oper.
{ T = Operation, L = 1, V = READ }
{ T = Channel, L = 15, V =
; L = 6 + 3*3 = 15
{ T= ChannelSelector,
L = 3, V: byte0 = portnum, byte1: channel type, byte2:
channelIndex }
{ T = TotalCW, L = 0 }
{ T = UncorrectedCW, L=0 }
{ T = CorrectableCW, L=0 }
}
{ T = Channel, L = 15, V =
{ T = ChannelSpecifier, L = 3, V: byte0 = portnum, byte1: channel type, byte2:
channelIndex }
{ T = TotalCW, L = 0 }
{ T = UncorrectedCW, L=0 }
{ T = CorrectableCW, L=0 }
}
}
06/15/15
CableLabs®
81
CM-SP-R-PHY-I01-150615
B.3.14.2
Data-Over-Cable Service Interface Specifications
RCP REX Message Normal Response Example
The message below represents a successful (no error) REX Response for stats counters for two upstream channels.
This is a response to the request outlined in Section B.3.14.1. As in previous examples, the outer envelope (GCP EDC
Normal Response) is not shown.
{ T = REX, L= 67, V=
; top-level “container” type
{ T = Sequence, L = 64, V =
{ T = Operation, L = 1, V = READ }
{ T = Channel, L = 27, V =
; L = 6 + 3*7 = 27
{ T= ChannelSelector,
L = 3, V: byte0 = portnum, byte1: channel type, byte2:
channelIndex }
{ T = TotalCW, L = 4, V = 32-bit counter}
{ T = UncorrectedCW, L= 4, V = 32-bit counter}
{ T = CorrectableCW, L=4, V = 32-bit counter}
}
{ T = Channel, L = 27, V =
{ T = ChannelSpecifier, L = 3, V: byte0 = portnum, byte1: channel type, byte2:
channelIndex }
{ T = TotalCW, L = 4, V = 32-bit counter}
{ T = UncorrectedCW, L= 4, V = 32-bit counter}
{ T = CorrectableCW, L=4, V = 32-bit counter}
}
}
}
B.3.14.3
RCP Rex Message Error Response Example
The example shown below represents a REX Response message for stats counters for two upstream channels. This is
a response to the request outlined in Section B.3.14.1. As in the previous examples, the outer envelope (GCP EDC
Normal Response) is not shown. The values of the response code (rspCode) are (TBD).
{ T = REX, L= 67, V =
{ T = Sequence, L = 64, V =
{ T = Operation, L = 1, V = READ-RSP }
{ T=RESP-CODE, L=1, V= rspCode }
;A Status TLV , only required on an error;
; if omitted, oper. was successful
{ T=ERROR-MSG, L=15, V="Unknown channel" } ; Optional RPD vendor specific msg for
the log
{ T = Channel, L = 15, V =
; L = 6 + 3*3 = 15
{ T= ChannelSelector,
L = 3, V: byte0 = portnum, byte1: channel type, byte2:
channelIndex }
{ T = TotalCW, L = 0, }
{ T = UncorrectedCW, L= 0}
{ T = CorrectableCW, L=0}
}
{ T = Channel, L = 15, V =
{ T = ChannelSpecifier, L = 3, V: byte0 = portnum, byte1: channel type, byte2:
channelIndex }
{ T = TotalCW, L = 0}
{ T = UncorrectedCW, L= 0}
{ T = CorrectableCW, L=0}
}
}
}
82
CableLabs®
06/15/15
Remote PHY Specification
B.3.14.4
CM-SP-R-PHY-I01-150615
An Example of an Embedded DOCSIS Message
The example shown below represents a REX Request Message, in which the CCAP-Core communicates to the RPD
the content of a DOCSIS message.
{ T = REX, L= nn + 13, V=
; top-level “container” type
{ T = Sequence, L = nn +13, V =
; nn is the length of the DOCSIS Message
{ T = Operation, L = 1, V = Write }
{ T = Channel, L = 3, V = nn + 15
{ T= ChannelSelector,
L = 3, V: byte0 = portnum; byte1: channel type, byte2:
channelIndex }
{T = DocsisMsg , L = nn, V = “Opaque Hex String forming a complete DOCSIS message”
}
}
}
}
06/15/15
CableLabs®
83
CM-SP-R-PHY-I01-150615
Annex C
Data-Over-Cable Service Interface Specifications
R-DEPI Extensions of R-UEPI Usage (Normative)
The R-UEPI control plane is the same as the R-DEPI control plane with additional AVPs. The R-DEPI control plane is
explained in [R-DEPI].
84
CableLabs®
06/15/15
Remote PHY Specification
Annex D
CM-SP-R-PHY-I01-150615
MPEG Stream Analysis (Normative)
The RPD MAY support MPEG Stream Analysis as described in this annex.
In order to validate that the MPEG stream served from the CCAP-Core does not have issues that will cause video
outages or other service impairments, the RPD will optionally be capable of performing tests on an MPEG stream to
verify its integrity. These checks are designed to detect video disruption and outages by detecting Packet Identifier
(PID) discontinuities and PID bitrate (or PID count) thresholds. The RPD monitors PIDs within both multi-program
and single-program transport streams (MPTS and SPTS) used to carry various MPEG system and control information,
video payloads, and audio payloads. PIDs monitored include the Program Association Table (PAT), the Program Map
Table (PMT), Program and System Information Protocol (PSIP), 0x1FFC carousel, and PIDs within a PMT program
such as video, audio, SCTE-35/Digital Program Insertion (DPI), and Enhaced TV Binary Interchange Format (EBIF).
If the RPD supports MPEG Stream Analysis, it MUST monitor MPEG synchronization by detecting transport stream
synchronization loss. A device synchronizes on a transport stream via the reception of correct sync bytes, which are
the 8 bits that precede the header of an MPEG packet (always 0x47). When the decoder first detects the sync byte, it
looks again for the next sync byte after 188 or 204 bytes in the stream. After finding three sync bytes in a row in this
pattern, synchronization has been established and packet boundaries are then known. However, if packets arrive with
incorrect sync bytes, synchronization loss occurs and the decoder again establishes MPEG synchronization. The RPD
will consider synchronization lost when two or more consecutive incorrect sync bytes are received.
Once the RPD has achieved MPEG synchronization, the following evaluations can be performed:
•
If the RPD supports MPEG Stream Analysis, it MUST be capable of reading the transport stream ID (TSID) from
the PAT. This value can be reported in enterprise MIBs. Note that a TSID is not available in DOCSIS streams.
•
If the RPD supports MPEG Stream Analysis, it MUST detect program PID discontinuity resulting in media loss.
•
If the RPD supports MPEG Stream Analysis, it MUST be capable of detecting the existence of the following
program PIDs in the transport stream:
•
•
•
•
•
•
PAT
PMT
Video
Audio
SCTE-35/DPI
EBIF
•
When loss of one of these PIDs is detected, it is expected that the RPD, using an SNMP trap or Syslog event, will
provide notification to the operator.
•
If the RPD supports MPEG Stream Analysis, it MUST be capable of detecting the existence of the mini-carousel
PID (0x1FEE) and reading the service group ID (SGID) from the PID.
In addition, the bit rates of certain PIDs can provide insight into the health of a given stream. For example, a too-low
bit rate could mean failure of the component providing the PID stream; a too-high bit rate could indicate an error
condition on that device. Either of these occurrences can cause service disruption. The rates of these can be monitored
via counters and a rate calculated on a time scale of minutes or several minutes to determine the health of the stream. If
the RPD supports MPEG Stream Analysis, it MUST monitor the PID bit rate of the following PIDs:
•
DOCSIS PID 0x1FFE
•
ATSC A65 PSIP base PID 0x1FFB
•
In-band DTA PIDs, including SI PID, 0x1FFC, and 0x1FF0
When an abnormal bit rate is detected, it is expected that the RPD, using an SNMP trap or Syslog event, will provide
notification to the operator.
06/15/15
CableLabs®
85
CM-SP-R-PHY-I01-150615
Annex E
Data-Over-Cable Service Interface Specifications
Certificate Hiearchy and Profiles (Normative)
This section describes the certificate format and extensions used by CableLabs certification authorities (CA) and
summarizes the fields of [X.509] version 3 certificates used for this specification. The CableLabs certificate PKI
hierarchy is shown below:
Figure 21 - Certificate Hiearchy
All certificates and CRLs described in this specification are signed with the RSA signature algorithm, using SHA-256
as the hash function. The RSA signature algorithm is described in PKCS #1 [RSA 1]; SHA-256 is described in [FIPS
180-4].
86
CableLabs®
06/15/15
Remote PHY Specification
E.1
CM-SP-R-PHY-I01-150615
CableLabs Root CA Certificate
The contents of the CableLabs Root CA Certificate is shown in Table 10.
Table 10 - CableLabs Root CA Certificate
Attribute Name
Settings
Version
v3
Serial number
Unique Positive Integer assigned by the CA
Issuer DN
c=US
o=CableLabs
ou=Root CA01
cn=CableLabs Root Certification Authority
Subject DN
c=US
o=CableLabs
ou=Root CA01
cn=CableLabs Root Certification Authority
Validity Period
50 yrs
Public Key Algorithm
Sha256WithRSAEncryption (1 2 840 113549 1 1 11)
Keysize
4096 bits
Parameters
NULL
Standard Extensions
OID
Include
Criticality
keyUsage
{id-ce
15}
X
TRUE
keyCertSign
Set
cRLSign
Set
basicConstraints
{id-ce
19}
X
TRUE
cA
subjectKeyIdentifier
Set
{id-ce
14}
X
FALSE
keyIdentifier
subjectAltName
Calculated per Method 1
{id-ce
17}
O
FALSE
directoryName
06/15/15
Value
Set by the issuing CA
CableLabs®
87
CM-SP-R-PHY-I01-150615
E.2
Data-Over-Cable Service Interface Specifications
CableLabs Device CA Certificate
The contents of the CableLabs Device CA Certificate is shown in Table 11.
Table 11 - CableLabs Device CA Certificate
Attribute Name
Settings
Version
v3
Serial number
Unique Positive Integer assigned by the CA
Issuer DN
c=US
o=CableLabs
ou=Root CA01
cn=CableLabs Root Certification Authority
Subject DN
c=US
o=CableLabs
ou=Device CA01
cn=CableLabs Device Certification Authority
Validity Period
35 yrs
Public Key Algorithm
Sha256WithRSAEncryption (1 2 840 113549 1 1 11)
Keysize
3072-bits
Parameters
NULL
Standard Extensions
OID
Include
Criticality
keyUsage
{id-ce
15}
X
TRUE
keyCertSign
Set
cRLSign
basicConstraints
Set
{id-ce
19}
X
TRUE
cA
Set
pathLenConstraint
0
subjectKeyIdentifier
{id-ce
14}
X
FALSE
keyIdentifier
authorityKeyIdentifier
Calculated per Method 1
{id-ce
35}
X
FALSE
keyIdentifier
subjectAltName
Calculated per Method 1
{id-ce
17}
O
FALSE
directoryName
88
Value
Set by the issuing CA for online
CAs
CableLabs®
06/15/15
Remote PHY Specification
E.3
CM-SP-R-PHY-I01-150615
RPD Certificate
The contents of the RPD Certificate is shown in Table 12.
Table 12 - RPD Certificate
Attribute Name
Settings
Version
v3
Serial number
Unique Positive Integer assigned by the CA
Issuer DN
c=US
o=CableLabs
ou=Device CA01
cn=CableLabs Device Certification Authority
Subject DN
c=<Country of Manufacturer>
o=<Company Name>
ou=DCA Remote Device Certificate
cn=<MAC Address>
Validity Period
20 yrs
Public Key Algorithm
Sha256WithRSAEncryption (1 2 840 113549 1 1 11)
Keysize
2048 bits
Parameters
NULL
Standard Extensions
OID
Include
Criticality
keyUsage
{id-ce
15}
X
TRUE
digitalSignature
Set
keyEncipherment
authorityKeyIdentifier
Value
Set
{id-ce
35}
X
FALSE
keyIdentifier
Calculated per Method 1
Values in angle brackets (<>) indicate that appropriate text as indicated below is present:
<Country of Manufacturer>: two-letter country code;
<Company Name>: name that identifies the company;
<MAC Address>: MAC address of the RPD.
The MAC Address is expressed as six pairs of hexadecimal digits separated by single colons (:), e.g.,
00:60:21:A5:0A:23. Hexadecimal digits greater than 9 are expressed as uppercase letters.
06/15/15
CableLabs®
89
CM-SP-R-PHY-I01-150615
E.4
Data-Over-Cable Service Interface Specifications
CableLabs Service Provider CA Certificate
The contents of the CableLabs Service Provider Certificate is shown in Table 13.
Table 13 - CableLabs Service Provider CA Certificate
Attribute Name
Settings
Version
v3
Serial number
Unique Positive Integer assigned by the CA
Issuer DN
c=US
o=CableLabs
ou=Root CA01
cn=CableLabs Root Certification Authority
Subject DN
c=US
o=CableLabs
ou=Service Provider CA01
cn=CableLabs Service Provider Certification Authority
Validity Period
35 yrs
Public Key Algorithm
Sha256WithRSAEncryption (1 2 840 113549 1 1 11)
Keysize
3072 bits
Parameters
NULL
Standard Extensions
OID
Include
Criticality
keyUsage
{id-ce
15}
X
TRUE
keyCertSign
Set
cRLSign
basicConstraints
Set
{id-ce
19}
X
TRUE
cA
Set
pathLenConstraint
0
subjectKeyIdentifier
{id-ce
14}
X
FALSE
keyIdentifier
authorityKeyIdentifier
Calculated per Method 1
{id-ce
35}
X
FALSE
keyIdentifier
subjectAltName
Calculated per Method 1
{id-ce
17}
O
FALSE
directoryName
90
Value
Set by the issuing CA for online
CAs
CableLabs®
06/15/15
Remote PHY Specification
E.5
CM-SP-R-PHY-I01-150615
CCAP-Core Device Certificate
The contents of the CCAP-Core Device Certificate is shown in Table 14.
Table 14 - CCAP-Core Device Certificate
Attribute Name
Settings
Version
v3
Serial number
Unique Positive Integer assigned by the CA
Issuer DN
c=US
o=CableLabs
ou= Service Provider CA01
cn= CableLabs Service Provider Certification Authority
Subject DN
c=<Country of Manufacturer>
o=<Company Name>
ou=DCA Headend/Hub Certificate
cn=<MAC Address>
Validity Period
2 yrs
Public Key Algorithm
Sha256WithRSAEncryption (1 2 840 113549 1 1 11)
Keysize
2048 bits
Parameters
NULL
Standard Extensions
OID
Include
Criticality
keyUsage
{id-ce
15}
X
TRUE
digitalSignature
Set
keyEncipherment
authorityKeyIdentifier
Value
Set
{id-ce
35}
X
FALSE
keyIdentifier
Calculated per Method 1
Values in angle brackets (<>) indicate that appropriate text as indicated below is present:
<Country of Manufacturer>: two-letter country code;
<Company Name>: name that identifies the company;
<MAC Address>: MAC address of the CCAP-Core Device.
The MAC Address is expressed as six pairs of hexadecimal digits separated by single colons (:), e.g.,
00:60:21:A5:0A:23. Hexadecimal digits greater than 9 are expressed as uppercase letters.
06/15/15
CableLabs®
91
CM-SP-R-PHY-I01-150615
E.6
Data-Over-Cable Service Interface Specifications
AAA Server Certificate
The contents of the AAA Server Certificate is shown in Table 15.
Table 15 - AAA Server Certificate
Attribute Name
Settings
Version
v3
Serial number
Unique Positive Integer assigned by the CA
Issuer DN
c=US
o=CableLabs
ou=Service Provider CA01
cn=CableLabs Service Provider Certification Authority
Subject DN
c=<Country>
o=<Company Name>
ou=AAA Server Certificate
cn=<server FQDN>
Validity Period
20 yrs
Public Key Algorithm
Sha256WithRSAEncryption (1 2 840 113549 1 1 11)
Keysize
2048 bits
Parameters
NULL
Standard Extensions
OID
Include
Criticality
keyUsage
{id-ce
15}
X
TRUE
digitalSignature
Set
keyEncipherment
authorityKeyIdentifier
Set
{id-ce
35}
X
FALSE
{id-ce
17}
X
FALSE
keyIdentifier
subjectAltName
Calculated per Method 1
dNSName
92
Value
<server FQDN>
CableLabs®
06/15/15
Remote PHY Specification
Appendix I
CM-SP-R-PHY-I01-150615
Plant Sweep in a Distributed Architecture (Informative)
Today, operators in HFC plants deploy test equipment that allows sweep tests to be performed, measuring plant
frequency response in the upstream and downstream direction. Traditionally, these have been closed, proprietary
systems with these characteristics:
•
In the downstream, proprietary equipment in the plant generates sweep signals that are measured by field test
equipment; a control channel between the headend equipment and test equipment controls how and when these
signals are generated.
•
In the upstream, the test equipment in the field generates signals that are measured by proprietary equipment in
the headend; a similar control channel between the test equipment and headend equipment is used to feed
measurements back to the test equipment so that adjustments can be made.
In a Remote PHY architecture, supporting the telemetry/control channel between the headend and the field test
equipment becomes a challenge. In a traditional architecture, the headend equipment is connected through the
combining network; this connection is eliminated in the R-PHY architecture. Other methods for performing sweep are
needed.
In this appendix, three alternatives to using currently available plant maintenance systems are discussed:
•
Using current transmitter and receiver technology, developed as part of the DOCSIS Proactive Network
Maintenance (PNM) toolset, to perform measurements;
•
Introducing modules to the R-PHY Node that perform the role of the headend test equipment;
•
Developing an API in the R-PHY Node that allows interaction with field test equipment.
I.1
Plant Sweep Using Transmitter and Receiver Capabilities
With the full-band capture capabilities introduced for DOCSIS 3.0 and 3.1 equipment, frequency response
measurements can be taken by either the CM or the R-PHY node receiver. Existing signals in the plant can be used in
the downstream for these measurements and the results of the measurements can be made available to test equipment
in the field via SNMP. To measure portions of the spectrum where no signals exist (for example, when evaluating
regions where services will be expanded for DOCSIS 3.1), the CCAP-Core can instruct the R-PHY node to generate
signals that can be measured by the CM.
In the upstream, existing signals can be measured and test modes on the CM can generate carrier signals that can be
measured at the R-PHY Node burst receiver. These measurements too can be exposed to field test equipment via
SNMP.
In addition, PNM enables symbol capture in both the upstream and downstream direction, allowing impairments to
also be detected in the time domain, rather than just the frequency domain.
Details on the PNM toolset can be found in the following DOCSIS 3.1 specifications: [CCAP-OSSI v3.1], [CM-OSSI
v3.1], [MULPI v3.1], and [PHY v3.1].
I.2
Hardware Module in the Node
Test equipment vendors may develop modules that will be deployed within a node that supports the R-PHY
architecture that performs the same function as the equipment that was previously deployed in the headend. Since the
module is located in the R-PHY Node, the same telemetry and control channels can be used. In this approach, the
sweep vendors can work with the node vendors to develop the sweep module and therefore the topic is not covered in
detail in this specification.
I.3
R-PHY Node API Support
In this approach, an API is developed by R-PHY Node and test equipment vendors that can be used by test equipment
to control the placement and configuration of signals in the RF spectrum. This API provides more control of sweep
carrier generation and access to measurements by the test equipment, without the need to support a specific hardware
module in the node, as described in the previous approach. Since the sweep signal itself is a CW signal, no additional
06/15/15
CableLabs®
93
CM-SP-R-PHY-I01-150615
Data-Over-Cable Service Interface Specifications
RF capability is required above what is defined in the R-PHY specifications (i.e., the ability to generate CW carriers at
any frequency and the ability to measure RF receive levels).
94
CableLabs®
06/15/15
Remote PHY Specification
Appendix II
CM-SP-R-PHY-I01-150615
Acknowledgements
On behalf of the cable industry and our member companies, CableLabs would like to thank the following individuals
for their contributions to the development of this specification:
Contributor
Company Affiliation
John T. Chapman
Cisco
Pawel Sowinski
Cisco
Gerry White
Cisco
Stuart Hoggan
CableLabs
Michael Patrick
Harmonic
On behalf of the cable industry and our member companies, CableLabs would like to thank the following individuals
for their contributions to the development of the technology and participation in the Remote PHY Working Group.
Contributor
Bill Powell
Company Affiliation
Alcatel-Lucent
Contributor
Nagesh Nandiraju
Company Affiliation
Comcast
Brian Kurtz
Altera
Saifur Rahman
Comcast
Carlton Lane
Analog
Jorge Salinger
Comcast
Linda Mazaheri
Analog
Joe Solomon
Comcast
Tom Ferreira
Arris
Douglas Will
Comcast
Steve Foley
Arris
Jeff Ford
Complex IQ
Anand Goenka
Arris
Al Garrett
Complex IQ
Jeff Howe
Arris
Ony Anglade
Cox Communications
Hari Nair
Arris
Mike Cooper
Cox Communications
Andrew Chagnon
Broadcom
Samir Parikh
Gainspeed Networks
Victor Hou
Broadcom
João Campos
Get
Niki Pantelias
Broadcom
Even Kristoffersen
Get
David Pullen
Broadcom
Adi Bonen
Harmonic
Stuart Hoggan
CableLabs
Mike Patrick
Harmonic
Huawei
Volker Leisse
CableLabs
Jim Chen
Karthik Sundaresan
CableLabs
Hesham ElBakoury
Huawei
Nikhil Tayal
CableLabs
Karl Moerder
Huawei
Jun Tian
CableLabs
Jack Moran
Huawei
Andrew Sundelin
CableLabs Consultant
Guangsheng Wu
Huawei
Naor Goldman
Capacicom
Phil Oakley
LGI
Dave Fox
Casa Systems
Stan Bochenek
Maxim Integrated
Maike Geng
Casa Systems
Ajay Kuckreja
Maxim Integrated
David Claussen
Charter
Len Dauphinee
MaxLinear
Nobo Akiya
Cisco
David Huang
MaxLinear
Alon Bernstein
Cisco
Louis Park
MaxLinear
Brian Bresnahan
Cisco
Sridhar Ramesh
MaxLinear
John T. Chapman
Cisco
Patrick Tierney
MaxLinear
Hang Jin
Cisco
Scott Walley
MaxLinear
Tong Liu
Cisco
Rei Brockett
Pace/Aurora
Carlos Pignataro
Cisco
Nasir Ansari
Rogers
Sangeeta Ramakrishnan
Cisco
George Hart
Rogers
John Ritchie
Cisco
Kevin Kwasny
Shaw
Pawel Sowinski
Cisco
Lee Johnson
ST Micro
Don Strausberger
Cisco
Paul Brooks
Time Warner Cable
06/15/15
CableLabs®
95
CM-SP-R-PHY-I01-150615
Contributor
Data-Over-Cable Service Interface Specifications
Company Affiliation
Contributor
Company Affiliation
Yi Tang
Cisco
Kirk Erichsen
Time Warner Cable
Bill Wall
Cisco
Colin Howlett
Vecima
Gerry White
Cisco
Douglas Johnson
Vecima
Philippe Perron
Cogeco
Faten Hijazi
Xilinx
John Bevilacqua
Comcast
Alex Luccisano
Xilinx
Additionally, CableLabs would like to thank the DCA MSO team for their continued support in driving the
specification development and the decision-making process.
Karthik Sundaresan, CableLabs
96
CableLabs®
06/15/15
Download PDF

advertising