Exploring weakness in Long Term Evolution (LTE) wireless standards Tien, Too Huseh

Exploring weakness in Long Term Evolution (LTE) wireless standards Tien, Too Huseh
Calhoun: The NPS Institutional Archive
Theses and Dissertations
Thesis Collection
2012-09
Exploring weakness in Long Term Evolution (LTE)
wireless standards
Tien, Too Huseh
Monterey, California: Naval Postgraduate School
http://hdl.handle.net/10945/48118
NAVAL
POSTGRADUATE
SCHOOL
MONTEREY, CALIFORNIA
THESIS
EXPLORING WEAKNESSES IN LONG TERM
EVOLUTION (LTE) WIRELESS STANDARDS
by
Too Huseh Tien
September 2012
Thesis Co-Advisors:
Weilian Su
Tri T. Ha
Approved for public release; distribution is unlimited
THIS PAGE INTENTIONALLY LEFT BLANK
REPORT DOCUMENTATION PAGE
Form Approved OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instruction,
searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send
comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to
Washington headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA
22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188) Washington DC 20503.
1. AGENCY USE ONLY (Leave blank)
2. REPORT DATE
3. REPORT TYPE AND DATES COVERED
September 2012
Master’s Thesis
4. TITLE AND SUBTITLE Exploring Weakness in Long Term Evolution (LTE)
5. FUNDING NUMBERS
Wireless Standards
6. AUTHOR(S) Too Huseh Tien
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
8. PERFORMING ORGANIZATION
Naval Postgraduate School
REPORT NUMBER
Monterey, CA 93943-5000
9. SPONSORING /MONITORING AGENCY NAME(S) AND ADDRESS(ES)
10. SPONSORING/MONITORING
N/A
AGENCY REPORT NUMBER
11. SUPPLEMENTARY NOTES The views expressed in this thesis are those of the author and do not reflect the official policy
or position of the Department of Defense or the U.S. Government. IRB Protocol number ______N/A______.
12a. DISTRIBUTION / AVAILABILITY STATEMENT
Approved for public release; distribution is unlimited
13. ABSTRACT (maximum 200 words)
12b. DISTRIBUTION CODE
A
The increasingly important role of Long Term Evolution (LTE) has increased security concerns among the service
provider and end users and made security of the network even more indispensable. In this thesis, the LTE
specifications are examined, and several security vulnerabilities of LTE mechanisms, in particular those that exist
within the Layer 2 protocol of the LTE network, are identified. Among these mechanisms, the power control
mechanism for LTE is further explored. The unprotected power control signal together with the Cell Radio Network
Temporary Identifier (CRNTI) can be exploited to trick the victim User Equipment (UE) to transmit at a much higher
than required power, which introduces significant inter-cell interference to adjacent base stations, evolved NodeB
(eNodeB). The ways that an attacker can maliciously manipulate the control field of the power control mechanism are
demonstrated. The effectiveness of such attack is evaluated with respect to the victim UEs and the adjacent eNodeBs.
The impacts include reduction of battery lifespan of victim UE to 33% of the original battery lifetime and reduction in
reverse channel signal-to-interference ratio (SIR) of adjacent eNodeB by 3.4 dB causing a decrease in throughput of
37%.
14. SUBJECT TERMS Long Term Evolution (LTE), Security, Power control, Vulnerabilities
17. SECURITY
CLASSIFICATION OF
REPORT
Unclassified
18. SECURITY
CLASSIFICATION OF THIS
PAGE
Unclassified
NSN 7540-01-280-5500
15. NUMBER OF
PAGES
103
16. PRICE CODE
19. SECURITY
20. LIMITATION OF
CLASSIFICATION OF
ABSTRACT
ABSTRACT
Unclassified
UU
Standard Form 298 (Rev. 2-89)
Prescribed by ANSI Std. 239-18
i
THIS PAGE INTENTIONALLY LEFT BLANK
ii
Approved for public release; distribution is unlimited
EXPLORING WEAKNESSES IN LONG TERM EVOLUTION (LTE) WIRELESS
STANDARDS
Too Huseh Tien
Defence Science & Technologies Agency
B.Eng (Electrical Engineering), Nanyang Technological University, 2005
Submitted in partial fulfillment of the
requirements for the degree of
MASTER OF SCIENCE IN ELECTRICAL ENGINEERING
from the
NAVAL POSTGRADUATE SCHOOL
September 2012
Author:
Too Huseh Tien
Approved by:
Weilian Su
Thesis Co-Advisor
Tri T. Ha
Thesis Co-Advisor
R. Clark Robertson
Chair, Department of Electrical and Computer Engineering
iii
THIS PAGE INTENTIONALLY LEFT BLANK
iv
ABSTRACT
The increasingly important role of Long Term Evolution (LTE) has increased security
concerns among the service provider and end users and made security of the network
even more indispensable. In this thesis, the LTE specifications are examined, and several
security vulnerabilities of LTE mechanisms, in particular those that exist within the Layer
2 protocol of the LTE network, are identified. Among these mechanisms, the power
control mechanism for LTE is further explored. The unprotected power control signal
together with the Cell Radio Network Temporary Identifier (CRNTI) can be exploited to
trick the victim User Equipment (UE) to transmit at a much higher than required power,
which introduces significant inter-cell interference to adjacent base stations, evolved
NodeB (eNodeB). The ways that an attacker can maliciously manipulate the control field
of the power control mechanism are demonstrated. The effectiveness of such attack is
evaluated with respect to the victim UEs and the adjacent eNodeBs. The impacts include
reduction of battery lifespan of victim UE to 33% of the original battery lifetime and
reduction in reverse channel signal-to-interference ratio (SIR) of adjacent eNodeB by 3.4
dB causing a decrease in throughput of 37%.
v
THIS PAGE INTENTIONALLY LEFT BLANK
vi
TABLE OF CONTENTS
I.
INTRODUCTION........................................................................................................1
A.
BACKGROUND ..............................................................................................1
B.
PROJECT OBJECTIVE .................................................................................7
C.
SCOPE OF THESIS ........................................................................................7
D.
APPROACH/STRUCTURE ...........................................................................8
II.
LITERATURE REVIEW ...........................................................................................9
A.
OVERVIEW .....................................................................................................9
B.
AUTHENTICATION PROTOCOL AND KEY MANAGEMENT
ENHANCEMENT IN LTE .............................................................................9
C.
WEAKNESS OF IP CONVERGED LTE NETWORK .............................10
D.
VULNERABILITIES OF NON-ACCESS-STRATUM (NAS) OF EUTRAN ...........................................................................................................10
E.
THREATS EXPLOITING LAYER 2 INFORMATION ...........................11
F.
INVESTIGATION OF LTE SPECIFICATION ........................................11
1.
Malicious Modification of Control PDU Type Reserve Field ........12
2.
Prioritized Retransmission of Traffic Data .....................................12
3.
Malicious Modification of Power Control Mechanism ..................12
III.
TECHNICAL BACKGROUND ...............................................................................13
A.
LTE TECHNOLOGY BASICS ....................................................................13
1.
OFDM .................................................................................................13
2.
OFDMA ..............................................................................................17
3.
SC-FDMA ...........................................................................................17
4.
MIMO Concept ..................................................................................19
5.
Generic Frame Structure ..................................................................21
6.
Physical Resource Block ....................................................................22
7.
Supportable Frequency Bands..........................................................23
B.
LTE NETWORK ARCHITECTURE OVERVIEW ..................................24
C.
NETWORK AND PROTOCOL ARCHITECTURE .................................28
1.
MAC [28] ............................................................................................29
2.
RLC [30] .............................................................................................31
3.
PDCP [31] ...........................................................................................33
4.
RRC [33] .............................................................................................34
D.
THREAT MODEL.........................................................................................34
E.
LTE SECURITY ............................................................................................34
1.
Control Plane Security ......................................................................35
2.
User Plane Security ............................................................................36
IV.
POTENTIAL WEAKNESS OF LTE SECURITY .................................................39
A.
CELL TYPE ...................................................................................................39
B.
INTERFERENCE ..........................................................................................39
C.
UPLINK POWER CONTROL .....................................................................41
1.
Closed Loop Power Control Mechanism (Normal) ........................43
vii
D.
E.
F.
V.
2.
Closed Loop Power Control Mechanism (Modified) ......................44
SCHEDULING GRANT ...............................................................................46
1.
Downlink Control Signaling..............................................................47
2.
Decoding and Search Space ..............................................................50
APPROACH ...................................................................................................52
1.
Stage 1– Acquisition of Cell Radio Network Temporary
Identifier (CRNTI) .............................................................................53
2.
Stage 2- Synchronization of Frame ..................................................53
3.
Stage 3- Message Injection ................................................................54
a.
Power Requirement for Message Injection ............................54
IMPACT .........................................................................................................58
1.
Depletion of Battery Power ...............................................................58
2.
Reduction of Reverse Channel SIR ..................................................59
CONCLUSIONS AND FUTURE WORK ...............................................................69
A.
CONCLUSIONS ............................................................................................69
B.
FUTURE WORK ...........................................................................................69
1.
Verification and Validation of Desired Received SINR .................69
2.
Investigation on Other Control Messages........................................70
APPENDIX A- MATLAB SIMULATIONS ........................................................................71
A.
CALCULATION AND PLOT OF FALSE SIGNAL TO
LEGITIMATE SIGNAL RATIO .................................................................71
B.
CALCULATION AND PLOT OF SIRAVE, NORMAL.....................................72
C.
CALCULATION AND PLOT OF SIRAVE, MAXIMUM ..................................72
LIST OF REFERENCES ......................................................................................................75
INITIAL DISTRIBUTION LIST .........................................................................................79
viii
LIST OF FIGURES
Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5.
Figure 6.
Figure 7.
Figure 8.
Figure 9.
Figure 10.
Figure 11.
Figure 12.
Figure 13.
Figure 14.
Figure 15.
Figure 16.
Figure 17.
Figure 18.
Figure 19.
Figure 20.
Figure 21.
Figure 22.
Figure 23.
Figure 24.
Figure 25.
Figure 26.
Figure 27.
Figure 28.
Figure 29.
Figure 30.
Figure 31.
Figure 32.
Figure 33.
Figure 34.
Figure 35.
Figure 36.
Figure 37.
Figure 38.
3GPP family technology evolution. From [2]....................................................2
3GPP LTE evolution country map. From [5]. ...................................................3
Representation of single-carrier transmission and OFDM in the frequency
domain. From [18]. ..........................................................................................13
Multipath-induced time delays result in ISI. From [19]. .................................14
Frequency-time representation of an OFDM signal. From [20]. .....................15
OFDM signal generation chain. From [20]......................................................16
Contrast between transmission schemes of OFDM and OFDMA. From
[18]. ..................................................................................................................17
SC-FDMA signal generation chain. From [18]. ..............................................18
Representation of OFDM and SC-FDMA signals. From [18]. ........................19
Principle of spatial multiplexing. From [20]....................................................20
Channel matrix H. From [21]. .........................................................................20
LTE frame structure type 1. From [22]. ...........................................................21
LTE frame structure type 2 (5 ms switch point periodicity). From [22]. ........22
Downlink resource grid. From [24] .................................................................23
High level architecture of LTE. After [25]. .....................................................25
Functional split between the E-UTRAN and EPC. From [26]. .......................26
User plane protocol stack. After [26]. ..............................................................27
Control plane protocol stack. After [26]. .........................................................28
Transmission of data in LTE downlink in time domain. From [27]. ...............28
Logical channels in LTE. After [28]. ...............................................................29
Transport channels in LTE. After [28]. ...........................................................30
Downlink mapping of logical to transport channels in LTE. From [29]. ........30
Uplink mapping of logical to transport channels in LTE. From [29]. .............31
Overview model of RLC sub-layer. From [30]. ..............................................32
RLC PDU structure. From [26]. ......................................................................33
Functional view of PDCP layer. From [30]. ....................................................33
Threat model for LTE network. After [25]. ....................................................35
Control plane layered security. After [34]. ......................................................36
User plane layered security. After [34] ............................................................37
Center cell antenna bearing orientation diagram. From [20]. ..........................40
Diagram of the network cell set-up with 120-degree directional antenna.
After [35]. ........................................................................................................40
Power control parameters transmitted from eNodeB to UE. ...........................42
Block diagram of steps involved in the closed loop power control
mechanism. ......................................................................................................43
Closed loop power control modified by adversary. .........................................44
Structure of MAC RAR. From [28]. ................................................................46
MAC PDU consisting of MAC header and MAC RARs. From [28]. ............46
Overview of PCFICH processing. From [29]. .................................................47
Overview of PHICH structure. From [29]. ......................................................48
ix
Figure 39.
Figure 40.
Figure 41.
Figure 42.
Figure 43.
Figure 44.
Figure 45.
Figure 46.
Figure 47.
Figure 48.
Downlink signal processing of the eNodeB. After [29]. .................................50
Search space of UEs in the control region. ......................................................52
Set-up position for MITM attack. ....................................................................56
Relation between the proximity of the attacker to the victim UE and the
required attacker’s transmitted power (n=4). ...................................................57
Battery lifespan of four LTE phones by applications. From [38]. ...................59
Reverse channel interference analysis for edge area. After [37]. ....................60
Signal-to-interference ratio for various combinations of UEs’ transmitted
power (UE1, UE2, UE3 and UE4 range from 10 mW to 200 mW). ...............64
Signal-to-interference ratio for first 100 combination of UEs’ transmitted
power to compute SIRAve,normal. ........................................................................65
Signal-to-interference ratio for various combinations of UEs’ transmitted
power (UE1, UE2 and UE3 are fixed at 200 mW). .........................................67
Throughput of a set of coding and modulation combination. From [39].........68
x
LIST OF TABLES
Table 1.
Table 2.
Table 3.
Table 4.
Table 5.
Table 6.
Table 7.
Table 8.
Table 9.
Table 10.
Table 11.
Table 12.
Comparison of parameters between LTE and its predecessors. From [3]. ........3
Major parameter for LTE Release 8. After [6]. .................................................6
User equipment categories for LTE Release 8. After [6]. .................................6
Uplink-downlink configuration for LTE frame structure type 2 [22]..............22
Resource block configuration for different channel bandwidths. From [24]...23
LTE operating band. From [24]. ......................................................................24
TPC commands with their corresponding values. From [36]. .........................45
Content for uplink scheduling grants. From [36].............................................45
DCI format with corresponding usage. From [29]...........................................49
Required transmitted power of attacker for various false-signal-tolegitimate-signal ratio. .....................................................................................57
Various input combinations of UEs’ transmitted power to compute
SIRAve,normal. ......................................................................................................64
Various input combinations of UEs’ transmitted power to compute
SIRAve,maximum. ....................................................................................................66
xi
THIS PAGE INTENTIONALLY LEFT BLANK
xii
LIST OF ACRONYMS AND ABBREVIATIONS
3GPP
Third Generation Partnership Program
4G
Fourth Generation
AKA
Authentication and Key Agreement
AM
Acknowledged Mode
AS
Access Stratum
BCCH
Broadcast Control Channel
BCH
Broadcast Channel
BPSK
Binary Phase Shift Keying
CCCH
Common Control Channel
CCE
Control Channel Element
CCI
Co-channel Interference
CFO
Carrier Frequency Offset
CN
Core Network
CP
Cyclic Prefix
CRNTI
Cell Radio Network Temporary Identifier
CRC
Cyclic Redundancy Check
CSG
Closed Subscriber Group
DCCH
Dedicated Control Channel
DCI
Downlink Control Message
DL-SCH
Downlink Shared Channel
DoS
Denial of Service
DRX
Discontinuous Reception
DTCH
Dedicated Traffic Channel
DwPTS
Downlink Pilot Timeslot
EAKA
Enhanced EPS-AKA
EAP-TLS
Extensible Authentication Protocol-Transport Layer Security
E-UTRA
Evolved Universal Terrestrial Radio Access
E-UTRAN
Evolved Universal Terrestrial Radio Access Network
EDGE
Enhanced Data rates for Global Evolution
eNodeB
Evolved NodeB (The base station in LTE system)
xiii
EPC
Evolved Packet Core
EPS
Evolved Packet System
ESIM
Enhanced Subscriber Identity Module
FDD
Frequency-Division Duplexing
FFT
Fast Fourier Transform
GP
Guard Period
GSA
Global mobile Suppliers Association
HARQ
Hybrid Automatic Repeat Request
HPSA
High Speed Packet Access
HSS
Home Subscriber Server
IFFT
Inverse Fast Fourier Transformation
ICI
Inter-Carrier interference
IMT
International Mobile Telecommunications
IP
Internet Protocol
IPSec
Internet Protocol Security
ISI
Inter-Symbol-Interference
LEAP
Lightweight Extensible Authentication Protocol
LTE
Long Term Evolution
LTE/SAE
Long Term Evolution/System Architecture Evolution
MAC
Medium Access Control
MCH
Multicast Channel
MBMS
Multimedia Broadcast Multicast Services
MBSFN
Multicast-Broadcast Single Frequency Network
MCCH
Multicast Control Channel
MCS
Modulation and Coding Scheme
MME
Mobility Management Entity
ME
Mobile Equipment
MIMO
Multiple-Input Multiple-Output
MITM
Man-In-The-Middle attack
MTCH
Multicast Traffic Channel
NAS
Non-Access Stratum
OFDM
Orthogonal Frequency-Division Multiplexing
xiv
OFDMA
Orthogonal Frequency-Division Multiple-Access
P-GW
Packet Data Network Gateway
PAP
Password Authentication Protocol
PAPR
Peak-to-Average Power Ratio
PBCH
Physical Broadcast Channel
PCFICH
Physical Control Format Indicator Channel
PCCH
Paging Control Channel
PCH
Paging Channel
PDCCH
Physical Downlink Common Control Channel
PDSCH
Physical Downlink Shared Channel
PDCP
Packet Data Convergence
PDU
Protocol Data Unit
PHICH
Physical Hybrid-ARQ Indicator Channel
PRACH
Physical Random Access Channel
PRB
Physical Resource Block
PUCCH
Physical Uplink Control Channel
PUSCH
Physical Uplink Shared Channel
QAM
Quadrature Amplitude Modulation
QoS
Quality of Service
QPSK
Quadrature Phase-Shift Keying
RACH
Random Access Channel
RAR
Random Access Response
REs
Resource Elements
REGs
Resource Element Groups
RF
Radio Frequency
RLC
Radio Link Control
RNTI
Radio Network Temporary Identifier
ROHC
Robust Header Compression
RRC
Radio Resource Control
S-GW
Serving Gateway
SAE
System Architecture Evolution
SC-FDMA
Single-Carrier FDMA
xv
SDU
Service Data Units
SIR
Signal-to-Interference Ratio
SINR
Signal-to-Interference and Noise Ratio
SNR
Signal-to-Noise Ratio
SKC
Session Keys Context
TDD
Time Division Duplexing
TDMA
Time-Division Multiple-Access
TPC
Transmit Power Control
UE
User Equipment
UEID
UE Identification
UL-SCH
Uplink Shared Channel
UMTS
Universal Mobile Telecommunications System
UpPTS
Uplink Pilot Timeslot
W-CDMA
Wideband Code-Division Multiple Access
WIMAX
Worldwide Interoperability for Microwave Access
xvi
EXECUTIVE SUMMARY
The rapid increase in data usage in mobile communication systems has led to the
development of the fourth generation (4G) Long Term Evolution (LTE) standard. LTE is
the current generation wireless data communications standard that is poised to dominate
mobile data connectivity in both the commercial and military arenas because of its very
high data rate capabilities. The increasingly important role of LTE has increased security
concerns among the service provider and end users and made security of the network
even more indispensable.
The literature related to the LTE network, in particular, that which is related to the
security issues, is reviewed and the security vulnerabilities identified in the available
literatures are discussed briefly. In this thesis, the LTE specifications are examined and
several security vulnerabilities of LTE mechanisms, in particular those that exist within
the Layer 2 protocol of the LTE network, are identified.
The identified potential
weaknesses include malicious modification of the control protocol data unit (PDU) type
reserve field, prioritized retransmission of traffic data, and malicious modification of the
power control mechanism.
In this thesis, the focus is on exploring the power control mechanism for LTE.
The objectives of power control are to improve the system capacity, coverage and user
experience, while at the same time reduce the power consumption of the User Equipment
(UE). Fundamentally, uplink power control for LTE is a combination of an open-loop
mechanism, where the UE transmit power depends on estimates of the downlink path
loss, and a closed loop mechanism, where the network directly controls the UE transmit
power by means of explicit transmitter power-control (TPC) commands transmitted in the
downlink. This closed loop mechanism is computed dynamically and updated from subframe to sub-frame. An adversary can inject false power-control commands to control the
UE transmit power, as shown in Figure 1.
xvii
Figure 1.
Modified power control mechanism.
The unprotected power control signal together with the Cell Radio Network
Temporary Identifier (CRNTI) can be exploited to change the intended behavior of the
UE. The CRNTI provides unique end User Equipment identification (UEID) at the cell
level and is assigned to the associated UE by the network during the initial establishment
of uplink synchronization. An adversary can exploit the fact that CRNTI is transmitted in
the clear and misuse it for malicious activities.
The ways that an attacker can maliciously manipulate the control field of the
power control mechanism are demonstrated in this thesis. The attacker acts as a
combination of base station, evolved NodeB (eNodeB), and the UE. Initially, the attacker
impersonates a UE and connects to the genuine eNodeB to obtain the cell–specific
reference signal. The attacker at a later stage presents itself as a bogus eNodeB and
generates false messages to the victim UE. The attacker can perform message injection
attack on the victim UE in three stages. Stage 1 involves the extraction of messages
between the victim UE and the eNodeB to obtain CRNTI. Stage 2 involves the
calculation of the timing advance to synchronize the false message frame to the victim
UE. Stage 3 involves the injection of false messages with the TPC field adjusted to the
xviii
designated value to change the behaviors of the victim UE. The correlation and graph for
the required power of the injected message for varying received false-signal-tolegitimate-signal is also derived.
The effectiveness of such an attack with respect to the victim UEs and adjacent
eNodeBs are evaluated. The impacts include reduction of battery lifespan of victim UE
and reduction in reverse channel signal-to-interference ratio (SIR) of adjacent eNodeB.
The interference generated by the victim UE in a 120-degree sectoring cell is
examined. A combination of these interferences creates a cascading effect on adjacent
eNodeBs, and the received SIR at the eNodeB is derived. From the derivation, it is
observed that SIR is dependent on the power transmitted by the UEs, and Matlab
simulation is performed to generate the average SIR. It is indicated in the simulation
results that the received SIR at the eNodeB decreases from a nominal value of 11.7 dB to
8.3dB when the interfering sources are transmitting at maximum power.
In general, a modulation and coding scheme (MCS) with a higher throughput
requires a higher SIR to operate in. The decrease in SIR leads to the adoption of an MCS
type with a lower throughput. The MCS is lowered from MCS-10, with a corresponding
maximum throughput of 3.2 bits per second per hertz, to MCS-8, with a corresponding
maximum throughput of 2.0 bits per second per hertz. The maximum throughput of the
legitimate UE is reduced by 37.5%.
.
xix
THIS PAGE INTENTIONALLY LEFT BLANK
xx
ACKNOWLEDGMENTS
I would like to thank my thesis advisors, Prof Tri Ha and Prof Weilian Su, for
their guidance and motivation to complete this thesis. I would also like to thank my
wonderful wife, Kim Hong, my family, and friends for their continuous support during
my study in Naval Postgraduate School.
xxi
THIS PAGE INTENTIONALLY LEFT BLANK
xxii
I.
A.
INTRODUCTION
BACKGROUND
“LTE is the next step in user experience, enhancing more demanding application
such as interactive TV, mobile video blogging, advanced gaming, and professional
services. Data rates are significantly higher. LTE supports a full [internet protocol] IPbased network and harmonization with other radio access technologies.” [1]
The rapid increase in data usage in mobile communication systems has led to the
development of fourth generation (4G) wireless technologies, which includes Long Term
Evolution (LTE) and Worldwide Interoperability for Microwave Access (WiMAX). LTE
is a standard developed by the Third Generation Partnership Program (3GPP) Long Term
Evolution/System
Architecture
Evolution
(LTE/SAE),
a
consortium
of
telecommunications associations formed in order to define communication standards, and
is specified in the 3GPP’s Release 8 document series, with minor enhancements
described in Release 9.
LTE belongs to the GSM path for mobile broadband and evolved after Enhanced
Data rates for Global Evolution (EDGE), Universal Mobile Telecommunications System
(UMTS), High Speed Packet Access (HPSA) and HSPA Evolution (HSPA+). The
evolutionary path is illustrated in Figure 1.
The first release of 3G provided by 3GPP in 2000 is known as “Release 99”. This
defines the wideband code-division multiple access (W-CDMA) and UMTS standards. In
2001, a new feature, “all-IP core network”, was added to Release 99, and it evolved to
Release 4. HPSA includes Release 5 and Release 6. Release 5 introduced the high speed
downlink packet access (HSDPA) in 2002 and Release 6 introduced the high speed
uplink packet access (HSUPA) and included more features like multimedia broadcast
multicast services (MBMS) and integration with wireless local area network (LAN) in
2005. Release 7 introduced HSPA+ in 2007 and primarily deals with the development of
specification like latency and quality of service (QoS) improvement and real time
applications.
1
Although the HSPA systems offer significant improvement in performance over
previous UMTS systems, their designs were limited by compatibility requirement in the
UMTS specification. In addition, with the emergence of packet-based mobile broadband
system like WiMAX, it is imperative for 3GPP to develop new standards and mobile
technologies to ensure competitiveness for the next decade and beyond in order to meet
the increasing demand of the network services, in terms of higher data rates, reduced
latency, improved system capacity and coverage. LTE/SAE proposes to fulfill these
requirements by using an IP converged architecture system which is able to work across
multiple access networks. Thus, LTE was first introduced in Release 8 in 2008, while
Release 9 is the LTE release with SAEs enhancement and the interoperability of LTE and
WiMAX.
The overall high level objective of LTE is to provide an extremely high
performance radio-access technology that provides full vehicular speed mobility and can
coexist with HSPA and other previous networks. With the scalable bandwidth
functionality of LTE, operators are able to migrate their networks and users from HSPA
to LTE over time with ease.
Figure 1.
3GPP family technology evolution. From [2].
LTE is able to provide unprecedented performance in terms of peak data rates,
delay, and spectrum efficiency to the network when compared with its predecessors. LTE
can provide up to 100 Mbps downlink data rate and up to 50 Mbps uplink data; this is
four times faster than previous HSPA+ data rates. The comparison of peak data rates and
other parameters between LTE and its predecessors are shown in Table 1.
2
Table 1.
Comparison of parameters between LTE and its predecessors. From [3].
WCDMA
(UMTS)
Maximum downlink speed (bps) 384 k
Maximum uplink speed (bps)
128 k
HSPA
(HSDPA/HSUPA)
14 M
5.7 M
HSPA+
28 M
11 M
100 M
50 M
Latency round trip time
(approximate)
3GPP releases
150 ms
Rel 99/4
100 ms
Rel 5/6
50ms
(Max)
Rel 7
~10 ms
Rel 8
Approximate years of initial roll
out
2003/4
2005/6 HSDPA
2007/8 HSUPA
2008/9
Access methodology
CDMA
CDMA
CDMA
LTE
OFDMA/SCFDMA
A LTE evolution update report researched and published by Global mobile
Suppliers Association (GSA) dated January 5, 2012, [4] confirms that 49 LTE operators
have already launched commercial services. These 49 LTE operators have launched LTE
networks services in 29 countries, which include Armenia, Australia, Austria, Bahrain,
Belarus, Brazil, Canada, Denmark, Estonia, Finland, Germany, Hong Kong, Hungary,
Japan, Kuwait, Latvia, Lithuania, Norway, Philippines, Poland, Puerto Rico, Saudi
Arabia, Singapore, South Korea, Sweden, UAE, Uruguay, USA, and Uzbekistan. The
countries with deployed LTE services are shaded in red in Figure 2.
Figure 2.
3GPP LTE evolution country map. From [5].
3
The GSA report also confirms that 285 operators in 93 countries have committed
to commercial LTE network deployments or are engaged in trials, technology testing or
studies. This includes the 49 commercial LTE network that are already launched, 117
deployments that are in progress or planned in 76 countries, and another 59 operators in
17 other countries that are engaged in LTE technology trials, tests or studies.
This report suggests that the operators around the world have strengthened their
commitment and investment in the LTE technology, and GSA forecast that there will be
119 commercial LTE networks in more than 50 countries by the end 2012.
The motivations for the growth of interest in LTE are as follows: continued
competitiveness of the 3G system, user demand for higher data rates and QoS, packet
switch optimized system, continued demand for reduced Capital and Operational
Expenditures (CAPEX and OPEX), low complexity, and avoidance of unnecessary
fragmentation of technologies for paired and unpaired band operation [6].
There are several key features of LTE discussed in [7]. These features are access
scheme, data rate, latency, mobility, spectrum allocation, frequency bands, scalable
bandwidth, cell size, supported users, internetworking with legacy network, packet
switched radio interface and support or Multicast-Broadcast Single Frequency Network
(MBSFN).
For the access scheme feature, LTE uses orthogonal frequency-division multiple
access (OFDMA) for the downlink and single carrier frequency-division multiple access
(SC-FDMA) for the uplink. The major parameters for LTE are shown in Table 2.
For the date rate feature, the peak download rates can support up to 299.6 Mbit/s
and upload rates up to 75.4 Mbit/s, depending on the User Equipment (UE) category.
Five different terminal classes have been defined from a voice centric class up to a highend terminal that supports peak data rates. The download and upload rates for respective
UE Categories are shown in Table 3.
For the latency feature, in optimal conditions, the data transfer latency is low at
sub-5 ms for small IP packets. This is lower for handover and connection set-up time than
with previous radio access technologies.
4
For mobility features, there is also an improved support for mobility, exemplified
by support for terminals moving at up to 350 km/h or 500 km/h depending on the
frequency band [8].
For the spectrum allocation feature, the LTE supports frequency-division
duplexing (FDD) and time-division duplexing (TDD) communication systems as well as
half-duplex FDD with the same radio access technology.
For the frequency bands feature, LTE supports all frequency bands currently used
by International Mobile Telecommunications (IMT) systems. For the scalable bandwidth
feature, LTE includes increased spectrum flexibility, with 1.4 MHz, 3 MHz, 5 MHz, 10
MHz, 15 MHz and 20 MHz wide cells standardized.
For the cell size feature, LTE supports cell sizes from tens of meters radius
(femto and picocells) up to 100 km radius macrocells. In the lower frequency bands to be
used in rural areas, 5 km is the optimal cell size, 30 km having reasonable performance,
and up to 100 km cell sizes supported with acceptable performance. In city and urban
areas, higher frequency bands (such as 2.6 GHz in the EU) are used to support high speed
mobile broadband. In this case, cell sizes may be 1 km or even less.
For the supported user feature, LTE supports at least 200 active clients in every 5
MHz cell [9]. For internetworking with legacy network, LTE supports the inter-operation
and co-existence with legacy standards.
For the MBSFN feature, LTE can deliver services such as Mobile TV using the
LTE infrastructure and is a competitor for DVB-H-based TV broadcast.
Several key enablers are required to achieve the aggressive performance targets of
the LTE. The identified key enablers for the LTE are orthogonal frequency-division
multiplexing (ofdm), multiple-input multiple-output (MIMO), and system architecture
evolution (SAE). [10]
The OFDM technology is an enabler in LTE because of its capability to transmit
at high data bandwidth efficiently while providing resilience to reflection and
interference. OFDMA is used in the downlink to achieve high peak data rates in high
5
spectrum bandwidth and SC-FDMA is used in the uplink because its small peak-toaverage power ratio; the more constant power enables high RF power amplifier efficiency
in mobile handsets, which is an important factor for battery power equipment.
Table 2.
Access Scheme
Bandwidth
Minimum TTI
Sub-carrier spacing
Cyclic prefix length
Modulation
Major parameter for LTE Release 8. After [6].
Uplink
DFTS-OFDM
Downlink OFDMA
1.4, 3, 5, 10, 15, 20 MHz
1 ms
15 kHz
4.7 µs
Short
16.7 µs
Long
QPSK, 16QAM, 64QAM
Single layer for Uplink per UE
Up to 4 layers for downlink per UE
MU-MIMO supported for uplink and downlink
Spatial multiplexing
Table 3.
User equipment categories for LTE Release 8. After [6].
Category
1
2
3
4
10
50
100
150
Peak rate
5
25
50
50
(Mbps)
Capability for physical functionalities
20MHz
RF bandwidth
QPSK,
16QAM,
64 QAM
Modulation Downlink
QPSK,
Uplink
16QAM
Multi-antenna
Assumed in performance requirements
2 RX diversity
Downlink
Uplink
2x2 MIMO
4x4 MIMO
Not
supported
5
300
75
Mandatory
Not supported
Mandatory
The MIMO technique is an enabler because one of the main problems
encountered by previous telecommunications systems was that multiple signals arose
from the many reflections that are encountered along the path. These signals would reach
the destination at different times and result in a disrupted waveform signal. With the
6
usage of MIMO, these additional signal paths can be used as an advantage to increase the
throughput. MIMO antenna technology enables ten times as many users per cell as
3GPP’s original WCDMA access technology [6].
Lastly, the SAE enables the system architecture to evolve in order to handle the
very high data rate and low latency requirements for 3G LTE. One of the significant
changes to the system architecture is that a number of the functions previously handled
by the core network have been transferred out to the periphery. This leads to a “flatter”
form of network architecture and allows direct routing of the data to the destination,
which in turn reduces the latency times.
With the superior features that LTE can provide, LTE is the next generation
wireless data communications standard that is poised to dominate mobile data
connectivity in both the commercial and military. In the commercial sector, a disruption
in service due to security reasons can jeopardize the reputation and reduce the revenue of
the service provider. In the military sector, the integrity and the timely transmission of
data are of upmost importance. Any compromise may result in failure of the mission.
Security is indispensable for secured communication between users and mobile
networks. The increasingly important role of LTE has brought about a number of security
concerns among the service provider and end users. The aim of this thesis is to provide a
comprehensive analysis on the potential weakness of the LTE protocol.
B.
PROJECT OBJECTIVE
Security within the LTE system has become extremely important to ensure
secured communication of the user terminals accessing network services. The security
and robustness of the LTE standard, especially those of its control channels in Layer 2,
namely, Radio Link Control (RLC), Medium Access Control (MAC) and Packet Data
Convergence (PDCP), need to be further examined.
C.
SCOPE OF THESIS
The scope of the thesis includes the review of the LTE protocol standards and the
assessment of existing threats to LTE system, followed by an exploration of methods of
7
hacking into and manipulating the control channel without the other party's knowledge.
This thesis research can serve as a starting point to protect, as well as to exploit, protocol
weaknesses in LTE and, thus, open exploitation space.
D.
APPROACH/STRUCTURE
The literature related to the LTE network is briefly discussed in Chapter II. In
particular, available literature related to security issues is reviewed and security
vulnerabilities are identified. The LTE specifications are examined, and several other
potential security weaknesses of the features and mechanisms, especially those related to
the control channels within the LTE network’s Layer 2 protocol, are identified.
In Chapter III, some of the important technical aspects of 3GPP are discussed.
Some basic technologies and methods employed in LTE, which include OFDM,
OFDMA, SC-FDMA and MIMO, are explored. A general overview of LTE architecture,
the different protocol layers and their interaction within the LTE network, followed by
the threat model and LTE’s security architecture are presented in Chapter III. Finally, the
details of sub-layer protocols, namely, RLC, MAC and PDCP within the LTE network’s
Layer 2 protocol, are elaborated on.
In Chapter IV, LTE’s power control mechanisms are explored and the unprotected
power control signal is exploited to conduct attacks on UEs and degrade their intended
services. The ways that an adversary can maliciously manipulate the power control
mechanism’s control field in order to sabotage the victim UE are demonstrated. This
chapter concludes by evaluating the impacts of the attack.
In Chapter V, the results of the thesis are summarized and the potential research
issues related to the security of LTE are discussed.
8
II.
A.
LITERATURE REVIEW
OVERVIEW
The majority of research related to LTE began in 2008, after the release of the
first LTE standard by 3 GPP. As the objective of the thesis is to examine the security and
robustness of the LTE standard, the literature review is related to materials discussing the
security aspect of LTE. A relatively small amount of research has been done on the
security of LTE, and only a limited number of security exploitations in LTE have been
discussed extensively in the published literature. There is, however, literature that serves
to provide a background and presents a tutorial overview of proposed security
mechanisms in Evolved Packet System (EPS), which lists some open security issues and
key threats in LTE at that time.
The threats discussed in [11] are the illegal access and usage of user’s and mobile
equipment’s (ME’)s identities, the tracking of user based on UE’s identity and signaling
messages, the illegal access and usage of keys used in security procedures, the malicious
modification of UE parameters to deny UE of normal services, the tampering with the
system information broadcasted to the Evolved Universal Terrestrial Radio Access
Network (E-UTRAN), the denial-of-service (DoS) to the UE, and the replaying attacks
which affect the integrity of data. These threats to the LTE network were not further
elaborated in [11].
B.
AUTHENTICATION
PROTOCOL
ENHANCEMENT IN LTE
AND
KEY
MANAGEMENT
Several researchers have done work to enhance the robustness of the security
protocol and mechanism in LTE.
The authors in [12] describe the LTE security architecture and mobility
procedures related to key management techniques in order to minimize the effects of a
possible key compromise in the access points. They go on to compare in detail LTE’s key
management security properties with the session keys context (SKC) concept. The
9
authors conclude that LTE could benefit from the SKC type of key management since
SKC concept is simpler and allows higher key distributor scalability, while the security
properties are quite similar.
The authors in [13] survey and compare three authentication protocols candidates:
Password Authentication Protocol (PAP), Lightweight Extensible Authentication
Protocol (LEAP) and Extensible Authentication Protocol-Transport Layer Security (EAPTLS) for LTE network. The conclusion is that PAP and LEAP are vulnerable to
dictionary attacks. EAP-TLS can provide reliable security performance, but has
considerable overhead.
The research in [14] examines the weaknesses and strength of the Authentication
and Key Agreement protocol (EPS-AKA) and identifies the protocol’s potential
weaknesses. A new authentication protocol, Enhanced EPS-AKA (EAKA), is proposed
which provides full (online) mutual entity authentication between ESIM (Enhanced
Subscriber Identity Module) and Home Subscriber Server (HSS) and removes the need
for delegated authentication.
C.
WEAKNESS OF IP CONVERGED LTE NETWORK
A survey of security threats conducted in [15] shows that the reason for the
unexpected service disruption and disclosure of information is the inherent weakness of
the converged Internet Protocol (IP) architecture of the LTE. The IP network is
susceptible to conventional attacks like IP address spoofing, user ID theft, theft of service
and DoS and intrusion attacks; these attacks are extended to the LTE network. In
addition, as mentioned in [16], the network is vulnerable to the known computer network
attack techniques such as man-in-the-middle (MITM), eavesdropping, Trojan, virus and
malware. Security vulnerabilities in the IP can jeopardize the entire IP converged LTE
network.
D.
VULNERABILITIES OF NON-ACCESS-STRATUM (NAS) OF E-UTRAN
The authors in [40] study the vulnerabilities of the Non-Access Stratum (NAS) of
E-UTRAN and illustrate attacks that exploit these vulnerabilities. The transmission of the
unprotected Radio Resource Control (RRC) messages and the transmission of
10
International Mobile Subscriber Identity (IMSI) in plain text without confidentiality and
integrity protection are discussed in the article. In addition, the CRNTI information in
Layer 1 provides attackers the opportunities to track an UE across cells. The exploitation
of these vulnerabilities allows the attackers to launch efficient and effective DoS attacks
on the eNodeB.
E.
THREATS EXPLOITING LAYER 2 INFORMATION
To the best of our knowledge, [17] is the only available reference that deals with
identifying threats and attacks by manipulating the information in MAC and RRC
signaling messages. The focus of [17] is on the security and privacy threats in radio
interface between eNodeB and the UE. There are two identified threats; the first threat is
the tracking of UE location based on the unique CRNTI, cell level measurement reports
or packet sequence numbers, and the second threat is the message insertion attack in
UE’s long discontinuous reception (DRX) period.
The long DRX period allows the UE to periodically switch off the processing
elements to save on the limited battery power and improves on power consumption’s
efficiency. However, this introduces extended delays when the UE needs to transmit or
receive data and may pose a security loophole for the system while the UE is “inactive”
during the long DRX period. The UE is vulnerable to attacks during this period; these
attacks includes false buffer status report attack which either steal bandwidth by changing
the packet scheduling behavior or changes the behavior of load balancing /admission
control algorithms in the eNodeB.
F.
INVESTIGATION OF LTE SPECIFICATION
Investigation of the LTE specifications revealed that there are vulnerabilities
within the LTE’s Layer 2 protocol. In this thesis, these vulnerabilities are identified and
the working principles are discussed briefly. The focus is on exploring LTE’s power
control mechanism. The ways to exploit the unprotected fields of the power control
message and attacks to the victim UE are detailed in Chapter IV.
11
Some of the potential vulnerabilities include the malicious modification of control
PDU type reserve field, the prioritized retransmission of traffic data, and the malicious
modification of power control mechanism. These are discussed in the following sections.
1.
Malicious Modification of Control PDU Type Reserve Field
The STATUS PDU is sent by the receiver to feedback on the status of the
received PDU. The control PDU type field is 3 bits and the STATUS PDU is indicated by
000, while 001-111 are reserved. PDUs with this reserved coding will be discarded by the
receiving entity for this release of the protocol (Release 10). The adversary can
maliciously adjust the control PDU type field to the reserved value and the recipient will
not be able to recognize and subsequently discard the STATUS PDU.
2.
Prioritized Retransmission of Traffic Data
The Radio Link Control (RLC) priority ruling [29] states that “the transmitting
side of an Acknowledged mode (AM) RLC entity shall prioritize retransmission of RLC
data Protocol Data Unit (PDUs) over transmission of new AM PDUs.” This implies that
when the transmitter receive a negative acknowledge on the previously PDU, it will
retransmit the missing PDU, instead of transmission of new data. This creates an
opportunity for the adversary to manipulate status update of the victim UE to negative
acknowledgement. This tricks the transmitter into continuously prioritizing and allocating
resource for the retransmission and reduces the chance of transmitting the legitimate data.
3.
Malicious Modification of Power Control Mechanism
The power control mechanism for LTE involves transmission of explicit TPC
control command to increase or decrease the transmission power of the UE. The
adversary can exploit the unprotected power control signal to conduct attacks on the UEs
and degrade their intended services. The impacts include depleting the limited battery
power of the UE at a faster rate, increasing interference to the neighboring cells.
12
III.
TECHNICAL BACKGROUND
Some of the important technical aspects of 3GPP LTE are discussed in this
chapter. The basic technologies and methods employed in LTE, which include OFDM,
OFDMA, SC-FDMA, and MIMO, are discussed in the following sections. In addition, an
overview of LTE architecture and the details of Layer 1 and Layer 2 protocols followed
by the LTE security and proposed threat model are presented. This aim of this chapter is
to provide the reader a preliminary background on LTE and aids him/her in
understanding the problem to be discussed in Chapter IV.
A.
LTE TECHNOLOGY BASICS
The LTE physical layer employs several advanced technologies to convey both
control and data information between the eNodeB and the UE. These techniques include
OFDM and MIMO data transmission.
1.
OFDM
Most cellular systems prior to LTE used single-carrier modulation schemes.
Although LTE uses OFDM instead of single-carrier modulation, it is imperative to
understand how the previous single-carrier systems dealt with multipath-induced channel
distortion and contrast that with OFDM systems. Graphical representations of singlecarrier transmission and OFDM in the frequency domain are shown and contrasted in
Figure 3.
Single Carrier
OFDM
Figure 3.
Representation of single-carrier transmission and OFDM in the frequency
domain. From [18].
13
In a communication system, delay spread refers to the amount of time delay at the
receiver from a signal travelling from the transmitter along different paths [19]. The delay
caused by multipath transmission can result in a received symbol from a delayed path to
“bleed” into a subsequence symbol that arrived at the receiver via the direct path. This
effect is known as inter-symbol interference (ISI) and is shown in Figure 4. In general,
the single-carrier system symbol time decreases as data rate increases, and it is possible
for ISI to spill into a second or third subsequent symbol at very high data rate.
Figure 4.
Multipath-induced time delays result in ISI. From [19].
Single-carrier systems usually compensate for channel distortion via time domain
equalization using either channel inversion or equalizers [19].
In channel inversion, a known sequence is transmitted over the channel prior to
sending actual information. As the original signal is known at the receiver, a channel
equalizer is able to determine the channel response and multiply the subsequent databearing signal by the inverse of the channel response to reverse the effects of multipath.
CDMA systems can employ equalizers to resolve the individual paths and then
combine digital copies of the received signal shifted in time to enhance the receiver
signal-to-noise ratio (SNR).
The implementation of channel equalizers is more complex as data rates increase.
The symbol times are shorter, and ISI is much more severe. The data rates of LTE is up
to 100 Mbps and delay spreads are about 17 µs [19]; thus, the approach of channel
equalization is unfeasible. Hence, OFDM is introduced, which eliminates ISI and greatly
simplifies the task of channel compensation.
LTE system employs OFDM as the downlink transmission scheme due to its
robustness against frequency selective-fading and narrowband interference. In OFDM,
14
the available bandwidth is spilt into multiple, narrow bandwidth sub-carriers and the data
is transmitted in parallel steams. Each sub-carrier is then independently modulated using
conventional modulation schemes such as quadrature phase-shift keying (QPSK), 16
quadrature amplitude modulation (QAM) or 64QAM and transmitted over the closely
spaced, orthogonal sub-carriers. A representation of the OFDM signal in the frequency
and time domain is shown in Figure 5. The problem of ISI is more severe as the data
transmission rate increases, and this problem occurs because the channel delay spread is
greater than the symbol period when the data is transmitted as a serial stream.
In OFDM, this problem is avoided by converting the data stream into multiple,
parallel sub-carriers. This conversion creates an OFDM symbol that is generally much
longer than the symbol on single-carrier systems and, thus, greater than the channel delay
spread. In Figure 5, the guard interval, which is the cyclic prefix (CP), is inserted prior to
the OFDM symbol in the time domain to eliminate ISI due to channel delay spread. The
use of narrow-band sub-carriers combined with the CP makes the transmission of OFDM
symbols inherently robust to the time dispersion on the channel and eliminates the need
for complex channel equalization on the receiver end. This property greatly simplifies the
processing required for the UE, which in turn reduces the terminal cost and the power
consumption.
Figure 5.
Frequency-time representation of an OFDM signal. From [20].
15
The generation of the OFDM signal is based on the inverse fast Fourier
transforms (IFFT), as illustrated in Figure 6. As shown in Figure 6, the IFFT converts N
frequency domain symbol streams to N complex time domain samples. These time
domain samples are then serialized to create the time domain signal.
Figure 6.
OFDM signal generation chain. From [20].
The superiority of OFDM to single-carrier systems in term of its ability to
eliminate ISI is discussed in previous section. OFDM, however, has two primary
weaknesses when compared to the single-carrier systems. OFDM is sensitive to carrier
frequency errors and has a large signal peak-to-average power ratio (PAPR).
One of the problems for OFDM is that it is sensitive to carrier frequency errors
due either to local oscillator offset or Doppler shifts [19]. Different reference frequencies
used in the transmitter and receiver can cause inter carrier interference (ICI) and result in
the loss of OFDM orthogonality. Also, the use of a cost effective local oscillator in the
UE may cause drifting of frequency and result in carrier frequency offset (CFO), which
may be greater than sub-carrier spacing.
Another disadvantage of OFDM is that it has a large signal PAPR. Amplitude
variations in the transmitted power of the single OFDM symbol are high because the
OFDM symbol is a combination of all of the sub-carriers, and the power these subcarriers can vary significantly. A high PAPR increases the dynamic range requirement of
the analog-to-digital and digital-to-analog converters and also reduces the efficiency of
the transmitter’s radio frequency (RF) power amplifier. The usage of a more expensive
transmitter capable of accommodating these requirements is often the remedy to the large
PAPR.
16
2.
OFDMA
In an OFDM transmission scheme, a single user receives all the sub-carriers at
one time. On the other hand, in an OFDMA transmission scheme, different users can
receive different subsets of sub-carriers simultaneously. Each user is allocated a specific
time-frequency resource, where data is transmitted over different sub-carriers over a
certain time period. The transmission scheme can be viewed in term of the time and
frequency domain. OFDM allocates resources to users in the time domain only, while
OFDMA allocates resources to users in both the time and frequency domains. A contrast
in the preceding transmission schemes is illustrated in Figure 7.
Figure 7.
Contrast between transmission schemes of OFDM and OFDMA. From [18].
3.
SC-FDMA
OFDMA is able to fulfill LTE’s high transmission data rate requirement while
eliminating ISI in the downlink as discussed in the previous section. The properties of
OFDMA signals, in particular the high PAPR, result in poorer uplink coverage. This
property makes it less favorable as an uplink transmission scheme for LTE.
SC-FDMA is selected as the LTE uplink transmission scheme since it can achieve
the benefits that OFDM brings to LTE because of similarities in the signal processing
17
properties of both transmission schemes. At the same time, SC-FDMA has a low PAPR.
This low PAPR characteristic is especially important for the design of a cost-effective
power amplifier for the UE.
The principle of discrete Fourier transform (DFT)-spread-OFDM is used to
generate the SC-FDMA signal as illustrated in Figure 8. The process is that an N-point
DFT is first input to a block of modulation data symbols in order to transform these
modulation symbols into frequency domain. The output of the transformed signal is then
mapped to the available sub-carriers, which then pass through an M-point IFFT operation
block. This is followed by parallel-to-serial conversion and the addition of CP. There are
two main schemes to implement the sub-carrier mapping, namely localized and
distributed. In a localized scheme, each user uses a set of adjacent sub-carriers to transmit
data. In a distributed scheme, each user uses sub-carriers that are spread across the entire
bandwidth.
Figure 8.
SC-FDMA signal generation chain. From [18].
In OFDM, each data symbol is modulated to each sub-carrier individually at a
given instant, and the digital modulation represents the amplitude of the respective subcarrier. Each sub-carrier of an OFDM signal carries information related to one specific
symbol. In contrast, in SC-FDMA, a linear combination of all the transmitted data
symbols at a given instant is modulated to a given sub-carrier, and all the transmitted subcarriers of the SC-FDMA signal carry a component of respective modulated data
symbols. Thus, each sub-carrier of the SC-FDMA signal carries information of all the
transmitted symbols. The representation of OFDMA and SC-FDMA signals are shown in
Figure 9.
18
Figure 9.
4.
Representation of OFDM and SC-FDMA signals. From [18].
MIMO Concept
MIMO technology is one of the key enablers for LTE to achieve the ambitious
requirement for high throughput and spectral efficiency through the use of multi-antenna
techniques at both the transmitter and receiver in the network. The improved performance
is achieved without additional bandwidth or increased transmission power. This is made
possible by dividing the same total transmission power over the multiple antennas to
achieve an array gain that improves the spectral efficiency (more bits per second per hertz
of bandwidth) or to achieve a diversity gain that improves the link reliability [19].
On a high level, LTE multi-antenna transmission can be divided into two modes,
namely spatial multiplexing and transmit diversity. Spatial multiplexing uses nonorthogonal MIMO codes to increase the bandwidth, while transmit diversity uses
orthogonal MIMO cods to increase power while preserving bandwidth. The use of one
MIMO mode or another depends on the radio channel condition.
Spatial multiplexing is a technique that allows transmission of multiple, different
data streams simultaneously on the same downlink resource block and is only possible if
the channel allows it [20]. These data streams can belong to a single user, which
significantly increases the peak rate of one user. These data streams can also belong to
different users, which increase the overall capacity. The principle of spatial multiplexing
is illustrated in Figure 10. As shown in Figure 10, spatial multiplexing exploits the
channel’s spatial dimension. The transmitted data stream go through a channel, which
19
consists of all NtNr paths between the Nt transmit antennas at the transmitter and the Nr
receive antennas at the receiver. This channel can be represented by the channel matrix
H, where hij represents the complex gain of the channel between the jth transmitter and
the ith receiver, as shown in Figure 11.
Figure 10.
Principle of spatial multiplexing. From [20].
Figure 11.
Channel matrix H. From [21].
On the other hand, transmit diversity can be used to increase the robustness of the
data transmission instead of increasing the data rate. Transmit diversity is a technique for
coherently adding the signals received from two transmit antennas. As the antennas are
physically separated, different channel impulse responses reduce the impact of deep
fading that occurs on each of the antenna, respectively, thereby enhancing the link
reliability.
20
5.
Generic Frame Structure
LTE physical layer transmission is deployable in two modes: frequency-division
duplexing (FDD) and time-division duplexing (TDD), each of which has its own frame
structure. The frame structure defines the frame slot and symbol in the time domain.
Although the uplink and downlink data transmission schemes are different, they share a
common frame structure.
Frame structure type 1 is defined for FDD mode, and the structure is as shown in
Figure 12. The LTE data transmission is segmented into frames which are 10 ms in
duration. Each frame consists of 10 sub-frames, and each sub-frame is further divided
into two slots period of 0.5 ms duration each.
Figure 12.
LTE frame structure type 1. From [22].
Frame structure type 2 is defined for TDD mode and is shown in Figure 13. The
LTE data transmission is also segmented into frames which are 10 ms in duration. Each
frame consists of two half frames. The half frame is further divided into four sub-frames
and a special sub-frame, or five sub-frames depending on the downlink to uplink switch
point periodicity.
The special sub-frames consist of three fields:
Downlink Pilot
Timeslot (DwPTS), Guard Period (GP) and Uplink Pilot Timeslot (UpPTS).
The frame structure of TDD can exist in seven different sub-frame format
configurations, with sub-frames 0 and 5 and DwPTS always reserved for downlink
transmission. The sub-frame that follows after the special sub-frame and UpPTS is
assigned to uplink transmission. The various uplink-downlink configurations are shown
21
in Table 4, where D denotes a sub-frame reserved for downlink transmission, U denotes a
sub-frame reserved for uplink transmission, and S denotes the special sub-frame.
Figure 13.
LTE frame structure type 2 (5 ms switch point periodicity). From [22].
Table 4.
6.
Uplink-downlink configuration for LTE frame structure type 2 [22].
Physical Resource Block
A physical resource block (PRB) is the smallest element of resource allocation
assigned by the base station scheduler [23]. LTE is a system with scalable bandwidth.
The current LTE specification defines six sets of supportable bandwidth from 1.4 MHz to
20 MHz with the corresponding PRBs required as shown in Table 5. Each PRB consists
of 12 consecutive sub-carriers of constant spacing of 15 kHz each, occupying a total
bandwidth of 180 kHz. A downlink slot consists of seven OFDM symbols when normal
CP is employed or six OFDM symbols when long CP is employed. A resource block
comprises of seven columns of OFDM symbols and 12 rows of sub-carriers, which
constitutes 84 resource elements, as shown in Figure 14.
22
Table 5.
Resource block configuration for different channel bandwidths. From [24].
Figure 14.
7.
Downlink resource grid. From [24]
Supportable Frequency Bands
The LTE specifications inherited the frequency bands defined for UMTS and
extended the list as shown in Table 6, where each E-UTRAN operating band with its
corresponding uplink and downlink operating band and duplex modes are displayed.
23
Table 6.
B.
LTE operating band. From [24].
LTE NETWORK ARCHITECTURE OVERVIEW
The high-level view of the LTE architecture network is shown and the interaction
of the various elements and interfaces are illustrated in Figure 15.
The architecture of the LTE is comprises of three main building blocks. They are
the UE, E-UTRAN and the Evolved Packet Core (EPC).
The UE is a mobile unit that allows a user to access network services, connecting
to the E-UTRAN via the radio interface.
The E-UTRAN consists of eNodeBs, which is another name for base stations, and
provides the user-plane (PDCP, RLC, MAC and physical layers) and control-plane
24
(RRC) protocol terminations towards the UE. The eNodeBs are typically interconnected
to each other by the X2 interface, enabling direct communication. The EUTRAN is
connected to the EPC by means of the S1 interface, and this connects the eNodeBs to the
mobility management entity (MME) and serving gateway (S-GW) elements.
The EPC is the core network in the LTE/System Architecture Evolution (SAE)
system and is responsible for overall control of the UE and establishment of the bearers,
which are the traffic flows between the UE and the Packet Data Network Gateway (PGW). The EPC is comprised of logical nodes, namely, P-GW, S-GW and MME.
EPC
E-UTRAN
UE
Figure 15.
High level architecture of LTE. After [25].
25
The functional split between the E-UTRAN and EPC is shown in Figure 16. The
yellow boxes in Figure 16 represent the logical nodes, white boxes represent the
functional entities of the control plane, and the blue boxes represent the radio protocol
layers.
Figure 16.
Functional split between the E-UTRAN and EPC. From [26].
The functions of the logical node eNodeB include radio resource management, IP
header compression and encryption of user data stream, the selection of an MME, the
routing of user plane data towards S-GW, the scheduling and transmission of paging
message, broadcast information and public warning system messages, the measurement
and measurement reporting configuration for mobility and scheduling, closed subscriber
group (CSG) handling that allows a permitted group of user to access a particular cell,
and a transport level packet marking in the uplink. [26]
The functions of the logical node MME are non-access stratum (NAS) signaling
(i.e., the signaling between the protocols that operates between UE and the Core Network
26
(CN)), NAS signaling security, access stratum (AS) security control, and inter-CN node
signaling for mobility between 3GPP access networks. [26]
The functions of the logical node S-GW include acting as the local mobility
anchor point for inter-eNodeB handover and mobility anchoring for inter-3GPP mobility,
E-UTRAN idle-mode downlink packet buffering and initiation of network triggered
service request procedure, lawful interception, and packet routing and forwarding.
The functions of the logical node P-GW consist of per-user based packet filtering,
lawful interception, UE IP address allocation, transport level packet marking in the uplink
and the downlink, and uplink and downlink service level charging, gating and rate
enforcement.
A comprehensive list of the functions offered by the logical nodes can be found in
3GPP 36.300. [26]
The user plane protocol stack consists of MAC, RLC and PDCP sub-layers that
are terminated at eNodeB as shown in Figure 17. The functions of these sub-layers are
discussed in the following sections. The control plane protocol stack is similar to user
plane protocol stack, with the exception of additional Radio Resource Control (RRC)
sub-layer terminated at eNodeB and NAS protocol terminated at MME, as shown in
Figure 18.
Figure 17.
User plane protocol stack. After [26].
27
Figure 18.
C.
Control plane protocol stack. After [26].
NETWORK AND PROTOCOL ARCHITECTURE
The relationships of the IP packet with the Protocol Data Unit (PDU) and Service
Data Units (SDU) at the respective layers are illustrated in Figure 19. In a data
transmission from the eNodeB to the UE, each protocol layer receives a SDU from higher
layer and appends the respective layer header to form and send the PDU to the lower
layer. In this study, the main focus is on the Layer 2 protocol. The PDCP, RLC and MAC
layers together constitute the Layer 2.
Figure 19.
Transmission of data in LTE downlink in time domain. From [27].
28
1.
MAC [28]
The MAC layer is mainly responsible for managing the mapping of logical
channels to the appropriate transport channels and the multiplexing and de
de-multiplexing
MAC SDUs between the physical and RLC layer. The various logical and transport
channels within LTE standard are illustrated in Figure 20 and Figure 21, respectively.
The supported mappings between
betwee these logical and transport
sport channels for the downlink
are displayed in Figure 22,
2 while those for the uplink are displayed in Figure 2
23. The
main transport channel for the downlink is DL
DL-SCH while that for uplink is UP
UP-SCH, as
shown in Figure 22 and 2
23, respectively. Other functions performed by MAC are the
hybrid automatic repeat
epeat request (HARQ) for retransmission function, scheduling
information reporting, and priority handling between UEs by means of dynamic
scheduling, priority handling between logical channels of one
ne UE, logical channel
prioritization, and transport format selection.
Logical Channels
Control Channel
Broadcast
Control Channel
(BCCH)
Traffic Channel
Common Control
Channel (CCCH)
Paging Control
Channel (PCCH)
Figure 20.
Multicast Control
Channel(MCCH)
Dedicated Traffic
Channel(DTCH)
Dedicated
Control
Channel(DCCH)
Logical channels in LTE. After [28].
29
Multicast Traffic
Channel (MTCH)
Transport
Channel
Downlink
Channels
Broadcast
Channel (BCH)
Paging Channel
(PCH)
Downlink
Shared Channel
(DL-SCH)
SCH)
Figure 21.
Figure 22.
Uplink Channels
Uplink Shared
Channel (ULSCH)
Random Access
Channel (RACH)
Multicast
Channel (MCH)
Transport channels in LTE. After [28].
Downlink mapping of logical
l
to transport channels
hannels in LTE. From [29].
The physical chan
channels defined in LTE include the physical broadcast channel
(PBCH), which carries
ies part of the system information required by the terminal in order to
30
access the network. The physical downlink shared channel (PDSCH) is used for unicast
transmission and for transmission of paging information. The physical downlink control
channel (PDCCH) is used for downlink control information, mainly scheduling decisions
and for scheduling grants enabling transmission on the physical uplink shared channel
(PUSCH). The physical hybrid-ARQ indicator channel (PHICH) carries the hybrid-ARQ
acknowledgement to indicate to the terminal whether a transport block should be
retransmitted or not. The physical control format indicator channel (PCFICH) is a
channel providing the terminals with information necessary to decode the set of
PDCCHs. The physical uplink shared channel (PUSCH) is the uplink counterpart to the
PDSCH. The physical uplink control channel (PUCCH) is used by the terminal to send
hybrid-ARQ acknowledgements, indicating to the eNodeB whether the downlink
transport block(s) was successfully received or not, to send channel-status reports aiding
downlink channel-dependent scheduling, and for requesting resources to transmit uplink
data upon. Finally, the physical random access channel (PRACH) is used for random
access. [29]
Figure 23.
2.
Uplink mapping of logical to transport channels in LTE. From [29].
RLC [30]
The RLC layer is the interface between the upper layers to the MAC layer as
illustrated in Figure 24. The RLC layer at the transmitter end is mainly responsible for
31
performing segmentation of RLC SDUs, where the IP packet is formatted to a
manageable size suitable for transmission at lower layer. The RLC layer at the receiver
end is responsible for the reassembly of RLC PDUs, where the PDU is formatted to fit
the MAC SDU. The RLC PDU structure is shown in Figure 25. RLC also performs the
reordering of RLC PDUs, duplicate detection and protocol error correction through
Automatic Repeat Request (ARQ).
The RLC layer provides three different modes: acknowledged, unacknowledged
and transparent for data transfer.
The functions of the acknowledged mode are as follows: the segmentation and
reassembly of RLC SDUs, the addition of RLC headers, the reliability in sequence
delivery service, and the suitability for carrying transmission control protocol traffic [31].
The functions of unacknowledged mode are as follows: the segmentation and
reassembly of RLC SDUs, the addition of RLC headers, no guarantee of delivery, and the
suitability for carrying streaming traffic [31].
In the transparent mode, there is no segmentation and reassembly of RLC SDUs,
no RLC headers added, no guarantee of delivery, but it is suitable for carrying voice [31].
Figure 24.
Overview model of RLC sub-layer. From [30].
32
Figure 25.
3.
RLC PDU structure. From [26].
PDCP [31]
The PDCP layer is mainly responsible for transfer, ciphering and deciphering of
user plane and control plane data. This layer also performs header compression and
decompression of IP data flows using the Robust Header Compression (ROHC)
protocols. Other functions performed by PDCP are integrity protection and integrity
verification of control plane data, maintenance of PDCP serial numbers, timer based
discard and duplicate discarding. The functional view of the PDCP layer is shown in
Figure 26.
Figure 26.
Functional view of PDCP layer. From [30].
33
4.
RRC [33]
The RRC layer is part of the LTE air interface control plane. This layer is
responsible for the broadcast of system information related to both the NAS and AS. It
performs RRC control such as paging, establishment, modification and release of RRC
connection, radio configuration and Quality-of-Service (QoS) control. Other functions
performed by the RRC layer are: inter- radio access technologies mobility, measurement
configuration and reporting, generic protocol error handling and support of selfconfiguration and self-optimization.
D.
THREAT MODEL
The proposed threat model for the LTE network is shown in Figure 27. In this
model, three elements are identified as being vulnerable to attack and are indicated by
the red arrows in Figure 27. These elements are the air interface between the UE and the
eNodeB, within the eNodeB, and the Internet protocol linkage between eNodeB and the
S-GW.
There is literature, as mentioned in Chapter II, that discusses the inherent
weakness of the IP network, and these weaknesses are susceptible to attacks from
Element 3. Thus, we will not discuss Element 3. Element 2, eNodeB, is typically
susceptible to physical attacks. We assume that the premises are secure and do not
discuss Element 2 either. The focus of this thesis is to study the possible attacks coming
from Element 1, which is the air interface between the UE and eNodeB. The objective is
to identify and exploit the unprotected control signaling between the UE and the eNodeB
and cause disruption or degradation of services to the UEs.
E.
LTE SECURITY
The LTE security architecture is designed to provide strong protection for control
signaling and the user data traffic exchanges between the different entities of the LTE.
The LTE architecture supports two distinct functions for the NAS and AS. The NAS
function comprises of end-to-end communication between the core network and the UE.
The AS function comprises of hop-by-hop communications between the network edges.
34
EPC
E-UTRAN
3
1
UE
Figure 27.
1.
2
Threat model for LTE network. After [25].
Control Plane Security
The LTE entities and the signals to secure the control plane interfaces are shown
in Figure 28. The control plane consists of NAS signaling between the UE and the
eNodeB, RCC signaling between the UE and the eNodeB, and S1-AP signaling between
the eNodeB and the MME. These signals are established between the entities and are
indicated in yellow boxes as illustrated in Figure 28. Encryption and integrity protection
of the NAS signaling is carried out in the NAS layer, while encryption and integrity
protection of the RRC signaling is performed at the PDCP layer. IP Security (IPSec)
tunneling is established between eNodeB and MME to carry the S1-AP signaling.
35
Figure 28.
2.
Control plane layered security. After [34].
User Plane Security
The LTE entities and mechanisms to secure user data traffic within the user plane
are shown in Figure 29. The user plane is protected by application protection between the
UE and the application server, user data protection between UE and the eNodeB and user
data protection between eNodeB and SAE-GW.
These protections are established
between the entities and are indicated in yellow boxes as illustrated in Figure 29.
Application providers are required to provide application layer protection between the
UE and the application server. User data protection between UE and the eNodeB is
provided using encryption and integrity protection at the PDCP layer, while user data
protection between eNodeB and SAE-GW is provided by established IPSec tunneling
[34].
36
Figure 29.
User plane layered security. After [34]
37
THIS PAGE INTENTIONALLY LEFT BLANK
38
IV.
POTENTIAL WEAKNESS OF LTE SECURITY
Three important metrics of a mobile network are data throughput, delay, and
power. The exploitation of the weaknesses in the protocol and the service mechanism that
causes service disruption or degradation on these three metrics are discussed in this
chapter. The disruption is typically achieved by exhausting the system’s limited
resources. In this paper, the LTE’s power control mechanisms are explored, and the
unprotected power control signal is exploited in order to conduct attacks on UEs and
degrade their intended services.
The background on the cell type structure used by the LTE network, the
interference experienced by UEs and eNodeB, and the power control mechanism utilized
by LTE are presented in the following sections. The ways that an adversary can
maliciously manipulate the control field of the power control mechanism to sabotage
victim UEs are demonstrated. The impacts of an attack on the victim UE, as well as the
neighboring eNodeB are evaluated at the end of the chapter.
A.
CELL TYPE
In this study, the LTE is assumed to operate in network cell with 120-degree
directional antennas, (i.e., each with three sectors per site/cell) with the base station in the
center of cell. This is in contrast with the classic network with omni-directional antennas,
which introduce more interference. The diagram of the 120-degree directional antenna
lobe for one cell sector is shown in Figure 30, while the diagram of a network cell set-up
with 120-degree directional antenna and adjacent cells is shown in Figure 31. In Figure
31, the different numbers represent the frequency channel band that users in the particular
sector are using.
B.
INTERFERENCE
The two types of interference considered include inter-cell and intra-cell
interference. Inter-cell interference is generated when the same carrier frequency is used
in adjacent cells. Intra-cell interference can arise in systems with non-orthogonal
channelization within the same cell. The main interference to the eNodeB is due to inter39
cell rather than intra-cell interference. The amount of interference to the neighboring UEs
within the same cell is effectively minimized in the ideal case since the LTE uplink is
orthogonal. However, there is a substantial amount of inter-cell interference to the
eNodeB from neighboring cells since adjacent cells have same frequencies assignments.
Generally, the closer a UE is to the neighboring cell, the stronger the generated
interference to that neighboring cell.
Figure 30.
Figure 31.
Center cell antenna bearing orientation diagram. From [20].
Diagram of the network cell set-up with 120-degree directional antenna. After
[35].
40
C.
UPLINK POWER CONTROL
Uplink power control for LTE refers to a set of tools by which the transmit power
for different uplink physical channels and signals are controlled to ensure that they are
received at the cell site with an appropriate power. The objectives of power control are to
improve the system capacity, coverage, and user experiences while at the same time
reduce the power consumption of the UE. In order to fulfill these objectives, power
control mechanisms are used to maximize the desired received power signal and to
minimize the amount of interference caused to the neighboring cells.
Fundamentally, the power control formula consists of two main portions. The first
part is computed according to the parameters signaled by the eNodeB. The second part is
computed dynamically and updated from sub-frame to sub-frame.
The overall closed loop power control for PUSCH transmission can be described
according to [36]. This transmitting power PT is set at the UE using the parameters
signaled by the eNodeB and is calculated as
PT = min{Pmax , P0 + α PLDL + 10 log10 ( M ) + ∆ mcs + δ }[dBm]
(1)
where Pmax is the maximum allowed transmit power of the particular UE class; Po is a cell
specific parameter that is broadcast as part of the system information, also seen as desired
received power; α is the path loss compensation factor; PLDL is the downlink path loss
estimated by the UE; M is the instantaneous bandwidth in terms of number of physical
resource block (PRB); ∆mcs is the different SINR required for the different modulation
schemes and coding rates; and δ is the explicit power control adjustment command.
Since Pmax is fixed, and the second term of the min function in Equation (1), i.e.,
Po+αPLDL+10log10(M)+ ∆mcs+δ, is variable, the UE transmit power is limited by Pmax.
In addition, the UE transmit power takes the lower value of the function in Equation (1).
To study the impact on the inter-cell interference to the eNodeB, some
assumptions and simplifications on the parameters used in Equation (1) are made. In
particular, Pmax is fixed at 23 dBm [24]; Po is assumed to be constant; α is assumed to be
1 with full compensation of path loss and is equal for all cells. In addition, the parameters
PLDL, M and ∆mcs are assumed constant, and finally, δ is maliciously set to its maximum
value.
41
To better appreciate the parameters involved in Equation (1), they are illustrated
in Figure 32. The parameters Po and α are signals that are broadcast at periodic intervals
of 160 ms. The parameter δ is the explicit power control command signal from the
eNodeB to the UEs at periodic intervals of 1 ms and constitutes to the dynamic part of the
power control equation. The typical use of the explicit power control command is to
compensate uplink multipath fading, which is not reflected in the downlink path loss. The
parameter PLDL is the path loss estimate calculated by the UE. This downlink path loss
can be estimated by measuring the received power of the downlink cell-specific reference
signals. The parameter Pmax is the maximum allowed transmit power of the UE.
Figure 32.
Power control parameters transmitted from eNodeB to UE.
42
1.
Closed Loop Power Control Mechanism (Normal)
Uplink power control for LTE is a combination of an open-loop mechanism,
where the UE transmit power depends on estimates of the downlink path loss, and closed
loop mechanisms, where the network can directly control the UE transmit power by
means of explicit power-control commands transmitted in the downlink.
The closed loop power control mechanism allows the UE to fine-tune the uplink
transmit power based on the transmitted closed loop correction value known as the
transmit power control (TPC) command. The TPC command is computed based on the
desired closed loop signal-to-interference and noise ratio (SINR) and the measured
(estimated) received SINR at the UE. When the received SINR is below the desired
SINR target, a TPC command is transmitted to the UE to request for an increase in the
transmitter power. If not, a decrease in transmitter power is requested. The computation
and the steps involved in the closed loop power control are illustrated in Figure 34. In
Figure 34, the boxes shaded in blue are actions performed by the eNodeB, while those in
brown are actions performed by the UE.
Figure 33.
Block diagram of steps involved in the closed loop power control mechanism.
43
The objectives of the closed loop power control to provide the required SINR are
to achieve an acceptable level of communication between the eNodeB and the UE and to
reduce the amount of interference received by the neighboring cells. At the same time,
power control aids in optimizing the limited battery power of the UE and achieves power
efficiency.
2.
Closed Loop Power Control Mechanism (Modified)
A malicious adversary can modify the TPC field to a large value during the
feedback loop transmission from eNodeB to UE as shown in Figure 34. In Figure 34, the
boxes shaded in red are actions performed by the malicious adversary. When the TPC
command field is adjusted to 7, corresponding to a value of 8 dB, this can increase the
transmit power to Pmax according to Equation (1) and trick the UE into transmitting power
at a higher power level. The respective TPC commands and values are shown in Table 7.
Figure 34.
Closed loop power control modified by adversary.
44
Table 7.
TPC commands with their corresponding values. From [36].
In the case of PUSCH transmission, the explicit power control command
controlling the term δ is included in the 20 bits uplink scheduling grants (UL grant). The
content of the 20 bits uplink scheduling grants is as shown in Table 8.
Table 8.
Content for uplink scheduling grants. From [36].
Field
Number of bits
Hopping flag
1 bit
Fixed size resource block assignment
10 bits
Truncated modulation and coding scheme
4 bits
TPC command for scheduled PUSCH
3 bits
UL delay
1 bit
CSI request
1 bit
The UL grant field is in the MAC Random Access Response (MAC RAR), which
also consists of three other fields: R, Timing Advance Command and Temporary CRNTI
as shown in Figure 35. A MAC PDU consists of a MAC header and zero or more MAC
RAR as shown in Figure 36.
45
TPC sub field
is within the
UL Grant
Figure 35.
Figure 36.
D.
Structure of MAC RAR. From [28].
MAC PDU consisting of MAC header and MAC RARs. From [28].
SCHEDULING GRANT
The uplink scheduling grant which includes the PUSCH resource indication,
transport format, and the command for power control of PUSCH uplink physical channel
is carried as Downlink Control Message (DCI) by the Physical Downlink Common
46
Control Channel (PDCCH). To aid in understanding how to modify the TPC field in the
uplink grant, it is imperative to study how this information is carried in the downlink
control channel.
1.
Downlink Control Signaling
Downlink control signaling is carried by three downlink control channels, namely,
the Physical Control Format Indicator Channel (PCFICH), the Physical Hybrid-ARQ
Indicator Channel (PHICH), and the Physical Downlink Common Control Channel
(PDCCH). Downlink control signaling is located at the start of each downlink sub-frame,
which spans up to the first three OFDM symbols.
The PCFICH indicates the size of the control region in term of the number of
OFDM symbols used for control signaling and is located in the first OFDM symbol of the
respective sub-frame. The PCFICH consists of two bits of information which correspond
to a control region size of one, two or three OFDM symbols. These two bits of
information are coded into a 32-bit codeword, scrambled with cell-and sub-framespecific scrambling code, QPSK-modulated for the transmission of 16 symbols. These 16
symbols are then mapped to four Resource Element Groups (REGs) where each REG
contains four Resource Elements (REs). These REGs are spread in frequency to achieve
good frequency diversity. The overall processing of PCFICH is illustrated in Figure 37.
The PCFICH-to-resource-element mapping depends on the cell identity to mitigate the
probability of inter-cell interference.
Figure 37.
Overview of PCFICH processing. From [29].
47
The PHICH consists of one bit of information and is used to acknowledge the
uplink data transmission. It is located in the first OFDM symbol of the respective subframe. The PHICH is spread on multiple REs to mitigate the power differences among
the REs and to provide sufficient energy for the transmission. In LTE, a structure is
adopted whereby several PHICHs are code multiplexed onto a set of REs as illustrated in
Figure 38. A PHICH group consists of eight PHICH (in case of normal cyclic) and is
transmitted on the same set of REs. As shown in Figure 39, the one bit of information for
the acknowledgement is repeated three times to form three information bits. It is then
modulated with binary phase-shift keying (BPSK) scheme on either the I or the Q branch,
followed by the spreading with a length-four orthogonal sequence. A composite signal
representing the group of PHICH is formed and scrambled. The twelve scrambled
symbols are then mapped to three REGs. These REGs are spread across frequency to
achieve good frequency diversity and to avoid inter-cell interference.
Figure 38.
Overview of PHICH structure. From [29].
The PDCCH is used to convey the DCI, including the downlink scheduling
assignments, uplink scheduling grants and power-control commands. The PDCCH is
mapped onto resource elements in one, two or three OFDM symbols in the first slot of a
sub-frame and is sent at every sub-frame interval. The message size of DCI depends on
the purpose of the control message. The DCI is defined into different DCI formats based
on sizes and usages and is summarized in Table 9.
48
Table 9.
DCI format with corresponding usage. From [29].
A PDCCH is transmitted on one or a group of several consecutive control channel
element (CCE), where a CCE is made up of nine REGs. The number of CCEs transmitted
(one, two, four, or eight) depends on the payload size of the DCI and the channel-coding
rate [29]. In PDCCH transmission, only those REGs which are not assigned to PCFICH
or PHICH are used, and multiple PDCCHs can be transmitted in a sub-frame.
The processing of the downlink signal is shown in Figure 39. First, the DCI
message and the RNTI are masked as a CRC attachment, which is then convolutionally
coded with a rate of 1/3 before producing the PDCCH bits. The PDCCHs bits to be
transmitted in a given sub-frame are then aggregated and scrambled by cell and subframe specific scrambling sequence, followed by QPSK modulation, interleaved and
cyclically shifted prior to PDCCH resource mapping.
49
PDCCH resource
Mapping
Figure 39.
2.
Downlink signal processing of the eNodeB. After [29].
Decoding and Search Space
Each PDCCH supports multiple formats, and the format used is unknown to the
UE. The UE is informed of the number of OFDM symbols within the control region of a
sub-frame but not explicitly informed of the detailed control channel structure. The
control region of the sub-frame comprises of PDCCHs for multiple UEs. The UE has to
monitor this particular area and blindly attempt to decode the control region in every subframe in order extract its own control information. The concept of UE Search Spaces
introduced in LTE enhances the UEs’ ability to decode the control channel region
50
efficiently. Instead of decoding the entire control channel region, a UE will only try to
decode CCEs within a pre-computed range known as the UE’s own Search Space.
An illustration on the mapping of the search space to the respective UEs in the
control region is shown in Figure 40. In this illustration, the size of the control region has
a length of three OFDM symbols. The specific starting location where the UE begins to
decode the CCEs corresponding to the PDCCHs is described in [36] and is calculated as
Z k = Yk mod  NCCE ,k / LPDCCH 
(2)
where Zk is the PDCCH search space staring location in sub-frame number k for CCE
aggregation level LPDCCH; NCCE is the number of CCEs in sub-frame number k; LPDCCH is
the CCE aggregation level, and sub-frame number k is an integer from 0 to 9.
The parameter Yk in Equation (2) is determined by
Yk = AYk −1 mod D
(3)
Yk −1 = 16 (UE _ ID ) + sub-frame number k
(4)
where Yk-1 is defined as
and A is 39822 while D is 65537.
This particular search space is determined by the sub-frame number and the UE’s
CRNTI [37]. The UE finds its PDCCH by monitoring a set of PDCCH candidates in
every sub-frame to extract downlink control information. Within the search space, the UE
de-masks each control candidate's CRC using its RNTI. If no CRC error is detected, the
UE considers it as a successful decoding attempt and reads the control information within
the successful candidate.
51
Figure 40.
E.
Search space of UEs in the control region.
APPROACH
The attacker acts as a combination of eNodeB and the UE. Initially, the attacker
impersonates a UE and connects to the genuine eNodeB to obtain the cell–specific
reference signal. At a later stage the attacker presents itself as bogus eNodeB and
generates false messages to the victim UE. The attacker can perform a message injection
attack on the victim UE, and this is to be performed in three stages. Stage 1 involves the
extraction of the messages between the victim UE and the eNodeB to obtain Cell Radio
Network Temporary Identifier (CRNTI). Stage 2 involves the calculation of the timing
52
advance in order to synchronize the false message frame to the victim UE. Stage 3
involves the injection of false messages with the TPC field sub field adjusted to the
designated value to change the behavior of the victim UE.
1.
Stage 1– Acquisition of Cell Radio Network Temporary Identifier
(CRNTI)
CRNTI provides unique end UE identification (UEID) at the cell level, and it is
assigned to the associated UE by the network during the initial establishment of uplink
synchronization. To achieve fast and flexible scheduling capability, the CRNTI is
transmitted with its scheduling information in the Layer 1 downlink control signal in
plain text [17]. Thus, the identity, CRNTI and its related resource allocation and other
Layer 1 control information are transmitted in the clear and are readable by anyone. The
vulnerabilities of the initial establishment of uplink synchronization provide the
opportunity for a man-in-the-middle attack machine to impersonate the legitimate UE and
the eNodeB. The adversary can exploit the fact that CRNTI is transmitted in the clear and
misuse it for malicious purposes.
As mentioned in the previous section, the eNodeB can perform CRC calculation
masked with the UE’s CRNTI on the control information, and the UE can de-mask the
control information using its own CRNTI within the search space. Thus, with the
captured CRNTI, the adversary can impersonate the eNodeB and inject false control
messages with adjusted control information field at predetermined timing and change the
intended behavior of the UE.
2.
Stage 2- Synchronization of Frame
Since OFDM systems are sensitive to time and frequency synchronization error
and in order to have the false message arrive at the UE simultaneously with the legitimate
message generated by the eNodeB, there is a need to acquire some form of
synchronization with the cell.
The adversary needs to perform cell search (similar to normal UE) to acquire
frequency and symbol synchronization to a cell, acquire frame timing of the cell that
determines the start of the downlink frame [29], and identify the cell–specific reference
53
signal. Based on the distance between the eNodeB and adversary UE and the distance
between the eNodeB and the victim UE, the adversary is able to calculate the time
difference of the two UEs upon reception of the same frame from the eNodeB. With the
calculated time difference, the adversary can determine the position of time slot relative
to the adversary’s UE when the first OFDM symbol (control region) of the frame reaches
the victim UE. With another round of calculation, the adversary can pre-determine the
timing advance required for the false message to be transmitted from the adversary’s UE
position. This enables the synchronized false message to arrive at the victim UE
simultaneously with the legitimate message.
3.
Stage 3- Message Injection
The adversary is able to determine the victim UE’s search space using the precaptured CRNTI and construct message to the particular UE’s search space and,
thereafter, inject the message according to the pre-determined timing. The injected false
message arriving at the victim UE will be of higher power than the message transmitted
from the legitimate eNodeB; thus, the legitimate message will be overwritten. Upon
receiving the message, the victim UE decodes the content of the control channel region
according to the search space and processes the information such as the scheduling
assignment and the scheduling grants.
a. Power Requirement for Message Injection
A typical set-up of a MITM attack is shown in Figure 41. In this set-up,
the position of the victim UE, the attacker and the eNodeB form an extended line. The
distance between the eNodeB and the victim UE and the transmitted power of the
eNodeB are denoted as d1 and PT,1, respectively. The distance between the malicious
attacker and the victim UE and the transmitter power of the malicious attacker are
denoted as d2 and PT,2, respectively. The received power at victim UE from eNodeB and
attacker are denoted as PR,1 and PR,2, respectively.
The received false-signal-to-legitimate-signal ratio SIRPr2/SIRPr1 at the
victim UE is derived in the following steps. The parameter SIRPr1 is calculated as
54
 PR ,1 

 I 
SIRPr1 = 
where I is the total co-channel interference received, and PR,1 is determined by
PR ,1 = GT ,1GR L−1 PT ,1
(5)
(6)
where GT,1 is the gain of the transmitter at eNodeB; GR is the gain of the receiver at the
victim UE; and L is the propagation loss. Substituting (6) into (5), we get
 GT ,1GR L−1 PT ,1 
SIRPr1 = 
.
I


(7)
The power law equation is determined as
L = β d1n
(8)
where β is a proportionality constant that is a function of the antenna heights of both
transmitter and receiver and the carrier frequency, and n is the path loss component
factor. Substituting (8) into (7), we get
 GT ,1GR β −1d1− n PT ,1 
SIRPr1 = 
 .

I


(9)
Similarly, for the attacker, the SIRPr2 is calculated as
 GT ,2GR β −1d 2 − n PT ,2 


I


SIRPr 2 = 
(10)
where GT,2 is the gain of the transmitter at attacker.
Assuming GT,1 = GT,2 and dividing (10) by (9), we get the received falsesignal-to-legitimate-signal ratio as
−n
SIRPr 2  d 2 PT ,2 
=  −n
.
SIRPr1  d1 PT ,1 
(11)
The power required for the malicious attacker to inject the false message is
dependent on the received SIR of the victim UE. This received SIR is in turn dependent
on both transmitters’ power, the distance between the transmitter and the receiver, and
55
the path loss component n as illustrated in (11). The set-up is assumed to be in lossy
environment where n is four. In order to effectively overwrite the legitimate message
from the eNodeB, the power of the injected false message must be significantly higher
than that of the former.
Figure 41.
Set-up position for MITM attack.
A graph for the required power of the injected message can be derived
based the distance ratio d2/d1 and the desired received false-signal-to-legitimate-signal
ratio PR,2/PR,1 at the victim UE as shown in Figure 42. The relationship between the
proximity of the attacker to the victim UE and the required attacker’s transmitted power
is shown in Figure 42.
An example is used to illustrate the required transmitted power of the
attacker based on the various false-signal-to-legitimate-signal ratios. In this example, the
56
victim UE is located 2 km from the eNodeB, while the attacker is 200 m from the UE.
This yields a distance ratio of 0.1. The eNodeB is transmitting at 30 W. From Figure 42,
we see that the required transmitted power of the attacker is only 0.02 times the amount
of the transmitted power of the eNodeB when the desired received SIR ratio at the victim
UE is 3 dB. This equates to only 0.6 W of power required for the attacker’s transmitter.
The required transmitted power of the attacker for the various remaining false-signal-tolegitimate-signal ratios is tabulated in Table 10. In general, the closer the attacker is to
the victim UE, the lower the power required to conduct the attack.
Figure 42.
Relation between the proximity of the attacker to the victim UE and the required
attacker’s transmitted power (n=4).
Table 10.
Required transmitted power of attacker for various false-signal-tolegitimate-signal ratio.
False-signal-tolegitimate-signal ratio
(dB)
3
6
10
13
Transmitted power of
attacker to eNodeB ratio
(times)
0.02
0.04
0.1
0.2
57
Transmitted
power of
attacker (Watt)
0.6
1.2
3
6
F.
IMPACT
The adversary’s action has two results. First, it depletes the limited battery power
of the UE at a faster rate and reduces the intended operation period. Second, it causes
interference to the neighboring cells. A combination of this interference from the
neighboring cells increases the interference perceived by the eNodeB and reduces the
desired SIR of the eNodeB. The decoding capability at the eNodeB is determined by the
SIR instead of the absolute received power. Thus, the increase of inferences of
neighboring UEs to the eNodeB reduces the SIR and changes the modulation and coding
scheme (MCS) to one, which lowers the maximum throughput. This in turn, restricts the
legitimate UEs to accessing their desired network services at a much lower data rate.
1.
Depletion of Battery Power
The battery lifespan of an end device is dependent on many parameters including
the device operation system, use applications, and user profiles. These applications in
turn determine the required bandwidth and the required transmit and receive power of the
end device. An approximate approach is used to explore the depletion rate of the battery
power of a device transmitting at 23 dBm, which is the maximum UE power specified in
[24]. Typically, a bandwidth demanding application like streaming video will require
higher transmit power for data transmission. As such, in this study, we assume the
estimated battery life of a phone continuously streaming video or browsing the web to
represent the battery life of the phone transmitting at 23 dBm. Also, we assume that the
estimated battery life of the phone performing an idle push email function to represent the
battery life of the transmitting at average power value. The estimated battery life of four
types of LTE phones by applications is plotted in Figure 43. For purposes of comparison,
we use the data of the Skyrocket phone to illustrate the UE depletion rate of battery
power for the various applications. From Figure 43, it is shown that Skyrocket will have
210 minutes of battery lifetime for streaming of video and will have 640 minutes of
battery lifetime for idle-push email. Analogously, a phone transmitting a maximum
power of 23 dBm has only 210 minutes of battery lifetime, which is a reduction of 430
58
minutes from a phone transmitting at nominal power. The phone battery lifetime is
reduced to 33% of the original battery lifetime when transmitting at maximum power.
Estimated Battery Life (Minutes)
Estimated battery life by application
1400
1200
1000
800
Steaming Video
600
Idle-Push Email
400
200
0
Skyrocket
Note
BIONIC
RAZR MAXX
Type of LTE phones
Figure 43.
2.
Battery lifespan of four LTE phones by applications. From [38].
Reduction of Reverse Channel SIR
The inter-cell interference condition is illustrated in Figure 44. The solid green
line indicates the desired transmit signal from the legitimate UE (UE4) located at the
corner of the outer cell of cell O to the eNodeB. Since the neighboring cell edge users
adopt sectoring, only cells D, E and F in the first tier which are facing the intended sector
(Channel 2) contribute to the co-channel interference (CCI). However, as none of these
three cells are using Channel 2, the interference comes from the second tier. In the second
tier, the only cells using Channel 2 and facing the intended sector are bottom cell A and
two cells B. The locations of UEs are designated as UE1, UE2 and UE3, respectively, as
shown in Figure 44. The solid red lines indicate the interference generated to the eNodeB
by UEs (UE1, UE2 and UE3) of adjacent cells.
59
Figure 44.
Reverse channel interference analysis for edge area. After [37].
The total co-channel interference I received by the eNodeB in cell O is given in
[41] and restated as follows
 GTGRPT , B   GTGRPT , A   GTGRPT , B ' 
I =
+
+

LB
LA
LB '

 
 

(12)
where GT is the gain of transmitter at the neighboring UE; PT,A, PT,B and PT,B’ are the
transmitted power from UE2, UE1, and UE3, respectively; and LA, LB and LB’ are the
propagation loss of transmitted power from UE2, UE1, and UE3, respectively. By
rearranging (12), we get
60
I = ( GTGR )( LB −1 PT , B + L A −1 PT , A + L B ' −1 PT , B ' ) .
(13)
The reverse channel SIR of the cell edge area (CE) for 120º-sectoring [41]
SIRCE,120º is defined as
 PR , O 
SIRCE ,120 = 

 I 
(14)
where PR,O is the received power from UE4 (legitimate user) and is calculated as
PR , O = GTGRLO −1 PT , O
(15)
where LO is the propagation path loss between UE4 and eNodeB, and PT,O is the
transmitted power of UE4. Substituting (13) and (14) into (15), we obtain


GTGRLO −1 PT , O

 ( GTGR ) ( LB −1 PT , B + LA −1 PT , A + LB '−1 PT , B ' ) 


SIRCE ,120 = 
(16)
where LB, LA, LB’ and Lo are the path losses for UE1, UE2, UE3 and UE4, respectively.
The LB value is calculated as
−n
LB = β B
−1
 21 
R  .

 2

(17)
where βB is a proportionality constant that is a function of the antenna heights of UE1,
and eNodeB, and R is the radius of the cell. The LA value is calculated as
LA = β A−1  3 2 3 R 
−n
(18)
where βA is a proportionality constant that is a function of the antenna heights of UE2 and
eNodeB. The LB’ value is calculated as
LB ' = β B '−1 (
7R)
−n
(19)
where βB’ is a proportionality constant that is a function of the antenna heights of UE3
and eNodeB. The LO value is calculated as
LO = β O −1 R − n
61
(20)
where βo is a proportionality constant that is a function of the antenna heights of UE4 and
eNodeB.
Substituting (17) to (20) into (16), we determine SIRCE,120º as
SIRCE ,120


=
  β B −1 



(
( β O −1 R − n ) PT , O
21 
R
2

−n
) P + (β
T, B
A
−1  3 3


R
 2

−n
)P
T, A
+ (β B ' (
−1
7R)
−n
)


.

PT , B '  

(21)
Assuming that βO=βA=βB=βB’ for the coverage area, we get the reverse channel
SIR of the CE as
SIRCE ,120


=
 
  



PT , O
.
−n
−n


−
n



21 
PT , B +  323  PT , A + ( 7 ) PT , B '  
2 



(22)
From (22), we observe that the SIR is dependent on the power transmitted by the
UEs and is independent of the cell radius. The average SIR can be computed to indicate
the average SIR experienced by the eNodeB and is given by
M
SIRAverage = ∑ pi SIRi
(23)
i =1
where SIRi represents the SIR experienced by the eNodeB computed by the respective
transmitted power combination of UE1, UE2, UE3 and UE4 as shown in Table 11. The
parameter pi represents the probability of that SIR value occurring, computed within the
specified transmitted power range, and M represents the number of transmitted power
combinations of UE1, UE2, UE3 and UE4.
Two types of average SIR, namely SIRAve,normal and SIRAve,maximum experienced by
the eNodeB are computed. The SIRAve,
normal
is the average SIR experienced by the
eNodeB based on the normal scenario where all the interfering power is random. On the
other hand, SIRAve, maximum is the average SIR experienced by the eNodeB based on the
extreme scenario where all the interfering power is at a maximum.
62
To formulate the value of the average SIRAve, normal experienced by eNodeB, it is
assumed that sampling is performed on the transmitted power of UE1, UE2, UE3 and
UE4. The transmitted powers can assume one of the twenty values, which range from 10
mW to 200 mW with steps of 10 mW. The SIRAve, normal is computed based on (22) and
(23), with various input combinations for the different transmitted power of UE1, UE2,
UE3 and UE4. There is a total of 204 =160,000 combinations of SIR with run-number 1
computed based on transmitted power of UE1, UE2, UE3 and UE4 being 10 mW and run
number 160,000 based on transmitted power of UE1, UE2, UE3 and UE4 being 200 mW
as shown in Table 11. A relatively lossy environment with n 4 is assumed in the
computation, and pi is 1/(number of combinations) where each combination of
transmitted power is equally likely to occur.
The results of the SIRAve,normal for the various combinations of the transmitted
power are simulated using Matlab code and are shown in Figure 45. The enlarged figure
for the first 100 combinations is shown in Figure 46.
Region 1 can be observed in Figure 45, while region 2 is illustrated in Figure 46,
which shows the first 100 data points of Figure 45. In region 1 of Figure 45, formed by
the first 8,000 combination runs, the SIR increases significantly to 10 dB at combination
run-number 21 as compared to the previous run. This occurs when the transmitted power
of UE3 is reset to 10 mW, while the transmitted power of UE2 is set to 20 mW. There is
an overall decrease in the interfering power from combination run-number 20 to 21. The
SIR generally follows a downward trend for this region until the transmitted power of
UE4 is set to 20 mW. At combination run-number 8,001, the SIR increases significantly
as the desired transmitted power of UE4 is set 20 mW as compared to the previous
10,000 combinations where the transmitted power of UE4 is at 10 mW. This pattern can
be observed for the subsequent 15,000 combination runs. Overall, the SIR increases as
the desired transmitted power of UE4 increases.
We observe that in region 2 of Figure 46, formed by the first 20 combination runs,
the SIR decreases as the interfering power of UE3 increases from 10 mW to 200 mW for
corresponding runs while the transmitted power of UE1, UE2, and UE4 remain at 10
mW. The SIR is inversely proportional to the interfering power.
63
Table 11.
Various input combinations of UEs’ transmitted power to compute
SIRAve,normal.
Run Combination
Number
1
2
3
4
5
…
80001
…
159996
159997
159998
159999
160000
Transmitted power (mW)
UE4
UE1
UE2
UE3
(Desired) (Interfering) (Interfering) (Interfering)
10
10
10
10
10
10
10
20
10
10
10
30
10
10
10
40
10
10
10
50
…
…
…
…
20
10
10
10
…
…
…
…
200
200
200
160
200
200
200
170
200
200
200
180
200
200
200
190
200
200
200
200
Region 1
Figure 45.
Signal-to-interference ratio for various combinations of UEs’ transmitted power
(UE1, UE2, UE3 and UE4 range from 10 mW to 200 mW).
64
Region 2
Figure 46.
Signal-to-interference ratio for first 100 combination of UEs’ transmitted power
to compute SIRAve,normal.
A Matlab simulation code is used to calculate the average SIRAve,normal
experienced by eNodeB according to (23) and the computed value is 11.7 dB.
When UE1, UE2 and UE3 are transmitting at maximum power of 200 mW each,
several assumptions were adopted to formulate the value of SIRAve, maximum experienced by
the eNodeB. First, it is assumed that the interfering transmitted power of UE1, UE2 and
UE3 was fixed at 200 mW. Second, the transmitted power of UE4 can assume one of the
20 values, which ranges from 10 mW to 200 mW with steps of 10 mW. The
SIRAve,
maximum
is computed based on the various combinations of different transmitted
powers of UE1, UE2, UE3 and UE4. There are a total of 20 combinations of SIR with
run-number 1 computed based on transmitted powers of UE1, UE2 and UE3 being 200
mW and UE4 being 10 mW, while run-number 20 is based on transmitted powers of
UE1, UE2, UE3 and UE4 being 200 mW as shown in Table 12. Third, a relatively lossy
environment with n 4 is assumed, and the parameter pi is 1/ (number of combinations),
where each combination of transmitted power is equally likely to occur. With these
assumptions, the value of the average SIRAve,
formulated.
65
maximum
experienced by eNodeB was
Table 12.
Various input combinations of UEs’ transmitted power to compute
SIRAve,maximum.
Combination
Number
1
2
3
4
5
…
16
17
18
19
20
Transmitted power (mW)
UE4
UE1
UE2
UE3
(Desired) (Interfering) (Interfering) (Interfering)
10
200
200
200
20
200
200
200
30
200
200
200
40
200
200
200
50
200
200
200
…
…
…
…
160
200
200
200
170
200
200
200
180
200
200
200
190
200
200
200
200
200
200
200
The results of the SIRAve, maximum are simulated using Matlab code, and they are
shown in Figure 47. In Figure 47, we observe that when the transmitted power of UE4 is
10 mW, the SIR is ˗2dB. This is because the overall CCI is higher than the desired
received signal of UE4, which resulted in the negative SIR. Overall, the SIR increases as
the desired transmitted power of UE4 increases. A Matlab simulation code is used to
calculate the SIRAve, maximum experienced by eNodeB, and the computed average value is
8.3dB.
The results show that there is a reduction in SIR of eNodeB by 3.4 dB, which is
calculated as 11.7 dB ˗ 8.3 dB, when the interfering transmitted power of UEs (UE1, UE2
and UE3) are fixed at maximum of 200 mW as compared to when the interfering
transmitted power of UEs varied from 10 mW to 200 mW. This lowers the MCS that can
be adapted by the victim UE with the eNodeB and reduces the data throughput
significantly.
The maximum throughput that can be achieved by a given MCS is the product of
the coding rate and the number of bits per modulation symbol [39]. Coding refers to
addition of redundant bits to the data bits and provides forward error correction on the
66
received bits, while coding rate is the proportion of the code bits to the data bits. The
order of modulation refers to the number of coded bits which can be transmitted per
modulation symbol.
Figure 47.
Signal-to-interference ratio for various combinations of UEs’ transmitted power
(UE1, UE2 and UE3 are fixed at 200 mW).
A graph of throughput for various MCS as a function of SINR is displayed in
Figure 48 [38]. Since mobile network is interference-limited where SNR has negligible
effect, SINR can be approximated to SIR [42]. Thus, Figure 48 is plotted against SINR
and can be used directly for our evaluation. A particular MCS requires a certain SIR in
order to operate with a suitably low bit error rate at the output. In general, a MCS with a
higher throughput requires a higher SIR. From Figure 48, we observe that to maximize
the throughput at around 11.7 dB, MCS-10 (16 QAM, R=4/5) corresponding to
throughput of 3.2 bits per second per hertz is the suitable MCS. When the SIR is reduced
to 8.3 dB, the MCS is MCS-8 (16 QAM, R=1/2), which corresponds to only 2.0 bits
second per hertz. Thus, the reduction in SIR of eNodeB from 11.7 dB to 8.3 dB will
decrease the maximum throughput of the UE by 37.5% from 3.2 bits per second per hertz
to 2.0 bits per second per hertz.
67
Figure 48.
Throughput of a set of coding and modulation combination. From [39].
68
V.
A.
CONCLUSIONS AND FUTURE WORK
CONCLUSIONS
This study consists of a comprehensive investigation of the LTE specifications
pertaining to Layer 2 protocols. As discussed in the literature review, the previous studies
on the security exploitation of Layer 2 protocols are not exhaustive. These studies are
limited to listing security vulnerabilities and do not elaborate further on the details and
the impact of respective threats.
This research has identified other potential vulnerabilities in the Layer 2 protocol
and demonstrated the potential of exploiting the unprotected power control message and
extracted CRNTI to change the intended behavior of the UEs. In particular, the victim UE
is tricked by a false message generated by a bogus eNodeB to transmit at a much higher
than required power, which introduced significant inter cell interference to the adjacent
eNodeB.
The impacts of the attack include depleting the limited battery power of the victim
UE at a much faster rate and reducing the reserve channel SIR of the eNodeB. The
intended phone battery lifetime is reduced to 33% of the original battery lifetime when
transmitting at maximum power. The simulation results show that there is a reduction in
reverse channel SIR of eNodeB by 3.4 dB, and this decreases the maximum possible
throughput of the UE by 37.5% from 3.2 bits per second per hertz to 2.0 bits per second
per hertz.
B.
FUTURE WORK
1.
Verification and Validation of Desired Received SINR
One important practical consideration that influences the amount of interference is
the changing environment where the LTE may be deployed, and the environment can
affect the path loss. This variation in environment was not simulated in this thesis. In
addition, the received SINR determines a range of MCS that can be adopted by the UE,
and the throughput varies for a different MCS at the same SINR. In this thesis, the
analysis is based only on the maximum possible throughput. The actual throughput loss
69
experienced by the UE when it operates with a SINR of 11.7 dB instead of 8.3 dB may
even be larger. Collection of the actual data can be used to validate and refine the results
of this research.
2.
Investigation on Other Control Messages
A thorough investigation can be conducted on RRC layer, in particular to the
unprotected RRC signaling. Some of these messages can be sent unprotected prior to
security activation, and some of the messages can be sent unprotected even after security
activation. The details on these messages can be found in 3GPP.36.331.
70
APPENDIX A- MATLAB SIMULATIONS
A.
CALCULATION AND PLOT OF FALSE SIGNAL TO LEGITIMATE
SIGNAL RATIO
clear all;
close all;
n = 4;
%Path loss exponent
P=0.01:0.01:0.2;
%From LTE (Rel-8), the maximum UE transmit power is
23dBm, that is, 200mW.
i=1;
improvement_3dB = 2;
improvement_6dB = 4;
improvement_10dB = 10;
improvement_13dB = 20;
for dist_ratio=0.05:0.025:1
power_ratio = dist_ratio ^ -n / improvement_3dB;
Display_3dB(i)= 1/power_ratio *100;
i=i+1;
end
i=1;
for dist_ratio=0.05:0.025:1
power_ratio = dist_ratio ^ -n / improvement_6dB;
Display_6dB(i)= 1/power_ratio *100;
i=i+1;
end
i=1;
for dist_ratio=0.05:0.025:1
power_ratio = dist_ratio ^ -n / improvement_10dB;
Display_10dB(i)= 1/power_ratio *100;
i=i+1;
end
i=1;
for dist_ratio=0.05:0.025:1
power_ratio = dist_ratio ^ -n / improvement_13dB;
Display_13dB(i)= 1/power_ratio *100;
i=i+1;
end
figure;
semilogy(0.05:0.025:1, Display_3dB, '-db');
hold on;
semilogy(0.05:0.025:1, Display_6dB , '-xr')
hold on;
semilogy(0.05:0.025:1, Display_10dB , '-og')
hold on;
71
semilogy(0.05:0.025:1, Display_13dB , 'magenta-')
grid on;
legend( 'Received ratio(Pr2/Pr1) of 3dB','Received ratio(Pr2/Pr1) of
6dB','Received ratio(Pr2/Pr1) of 10dB','Received ratio(Pr2/Pr1) of
13dB');
xlabel('Ratio of distance (D2/D1)')
ylabel('Ratio of transmitted power (Pt2/Pt1)(%)')
title('Plot of relationship between ratio of distance vs ratio of
transmitted power for various received power ratio')
B.
CALCULATION AND PLOT OF SIRAVE, NORMAL
clear all;
close all;
n = 4;
%Path loss exponent
P=0.01:0.01:0.2;
%From LTE (Rel-8), the maximum UE transmit power is
23dBm, that is, 200mW.
i=1;
for Pto=0.01:0.01:0.2
for Pta=0.01:0.01:0.2
for Ptb=0.01:0.01:0.2
for Ptbl=0.01:0.01:0.2
SIR_ceu1_rev = (Pto/(((((sqrt(21))/2)^-n)*Ptbl) +
((((3/2)*(sqrt(3)))^-n)*Pta) + (((sqrt(7))^-n)*Ptb)));
Display(i)= 10*log10(SIR_ceu1_rev);
i=i+1;
end
end
end
end
figure, plot(Display, 'blueo', 'Markersize', 1);
xlabel('Combination run number')
ylabel('Signal-to-interference ratio (dB)')
title('Plot of Signal-to-interference ratio for various combinations of
UEs transmitted power')
C.
CALCULATION AND PLOT OF SIRAVE, MAXIMUM
clear all;
close all;
n = 4;
%Path loss exponent
P=0.01:0.01:0.2;
%From LTE (Rel-8), the maximum UE transmit power is
23dBm, that is, 200mW.
72
%Interfering source at max of 23dBm, ie, 200mW
Pta=0.2;
Ptb=0.2;
Ptbl=0.2;
Z1_rev=0;
i=1;
for Pto=0.01:0.01:0.2
SIR_ceu1_rev = (Pto/(((((sqrt(21))/2)^-n)*Ptbl) +
((((3/2)*(sqrt(3)))^-n)*Pta) + (((sqrt(7))^-n)*Ptb)));
Display(i)= 10*log10(SIR_ceu1_rev);
i=i+1;
end
figure, plot(10:10:200,Display);
xlabel('Transmitted power of UE4 (mW)')
ylabel('Signal-to-interference ratio (dB)')
title('Plot of Signal-to-interference ratio for various transmitted
power of UE4 (Desired)')
73
THIS PAGE INTENTIONALLY LEFT BLANK
74
LIST OF REFERENCES
[1]
Global mobile Suppliers Association (2012, January). GSM/3G
Market/Technology Update [Online]. Available: http://www.gsacom.com
[2]
4Gamericas. Long Term Evolution [Online] Available:
http://www.4gamericas.org/index.cfm?fuseaction=page&sectionid=249
[3]
Radio Electronics. 3GPP Long Term Evolution [Online]. Available:
http://www.radio electronics.com/info/cellulartelecomms/lte-long-termevolution/3g-lte-basics.php
[4]
Global mobile Suppliers Association (January 2012). LTE report [Online].
Available: http://www.gsacom.com/news/gsa_344.php4
Wikipedia (January 2012). 3GPP LTE Evolution Country Map: Adoption of LTE
as of January 5, 2012 [Online]. Available:
http://en.wikipedia.org/wiki/File:3GPP_Long_Term_Evolution_Country_Map.sv
g]
3GPP (2012). LTE [Online]. Available: http://www.3gpp.com/LTE
[5]
[6]
[7]
Wikipedia (January 2012). LTE (Telecommunication) [Online]. Available:
http://en.wikipedia.org/wiki/3GPP_long_term_evolution
[8]
S.T. Baker, “Introduction and Background” in LTE – The UMTS Long Term
Evolution; From Theory to Practice, Wiley, pp 1-20, 2009.
[9]
LTE World. Evolution of LTE. [Online] Available: http://lteworld.org/blog/lteadvanced-evolution-lte
[10]
Radio Electronics. 3GPP Long Term Evolution [Online]. Available:
http://www.radioelectronics.com/info/cellulartelecomms/lte-long-termevolution/3g-lte-basics.php
[11]
C.B. Sankaran, “Network Access Security in Next-Generation 3GPP Systems: A
Tutorial,” IEEE Commun. Mag., vol. 47, no. 2, Feb. 2009.
[12]
D. Forsberg, “LTE Key Management Analysis with Session Keys Context,”
ELSEVIER Computer Communication, vol. 33, no. 16, Oct. 2010.
[13]
R.Narmadha, and Dr.S.Malarkkan, “Review of security Analysis in LTE and
WIMAX Environment” IEEE International Conference on Computational
Intelligence and Computing Research, 2010.
[14]
G. M. Køien, “Mutual Entity Authentication for LTE,” IEEE IWCMC, Jul. 2011.
[15]
Y. Park, et al., “A Survey of Security Threats on 4G Networks,” IEEE
GLOBECOM Workshop on Security and Privacy in 4G Networks, Nov. 2007.
75
[16]
Beaumont, J.-and Doucet, G, "Threats and Vulnerabilities of Next Generation
Satellite Personal Communications Systems: A Defence Perspective," Globecom
Workshops, IEEE, pp.1-5, 26-30 Nov. 2007.
[17]
D. Forsberg, et al., “Enhancing Security and Privacy in 3GPP E-UTRAN Radio
Interface,” IEEE PIMRC, Sep. 2007.
[18]
ROHDE & SCHWARZ (April 2009). LTE technology and LTE test;a deskside
chat [Online]. Available: http://www.telecom-cloud.net/wp
content/uploads/2010/11/Rohde-and-Schwarz-LTE-Tutorial.pdf
[19]
Wikipedia (2012 January). MIMO [Online]. Available:
http://en.wikipedia.org/wiki/MIMO
[20]
Technical Specification Group Radio Access Network; Feasibility Study for
Orthogonal Frequency Division Multiplexing (OFDM) for UTRAN enhancement,
3GPP TR 25.892.
[21]
4GWirelessjob(2012). LTE MIMO Concepts [Online]. Available:
http://4gwirelessjobs.com/articles/article-detail.php?LTE-MIMOConcepts&Arid=MTQz&Auid=OTY=
[22]
Evolved Universal Terrestrial Radio Access (E-UTRA); Physical channels and
modulation, 3GPP TS 36.211, 2011.
[23]
J. Zyren, “Overview of the 3GPP Long Term Evolution Physical Layer”,
3GPPEVOLUTIONWP, 2007.
[24]
Evolved Universal Terrestrial Radio Access (E-UTRA); User Equipment radio
transmission and reception, 3GPP TS 36.101, 2012.
[25]
Magnus Lindstrom,” LTE-Advanced Radio layer 2 and RRC aspect”. 3GPP LTEAdvanced Evaluation Workshop, 2009.
[26]
Evolved Universal Terrestrial Radio Access (E-UTRA) and Evolved Universal
Terrestrial Radio Access Network (E-UTRAN); Overall description; Stage 2,
3GPP TS 36.300, 2012.
[27]
Freescale, “Long Term Evolution Protocol Overview” LTEPTCLOVWWP, 2008.
[28]
Evolved Universal Terrestrial Radio Access (E-UTRA); Medium Access Control
(MAC) protocol specification, 3GPP TS 36.321, 2012.
[29]
E.Dahlman, S.Parkvall, J.Slold, P.Beming et al., “3G Evolution HSPA and LTE
for Mobile Broadband”, 2nd ed. Academic Press, 2008.
76
[30]
Evolved Universal Terrestrial Radio Access (E-UTRA); Radio Link Control
(RLC) protocol specification, 3GPP TS 36.322, 2011.
[31]
EventHelix (2009). 3GPP LTE Radio Link Control sub-layer [Online]. Available:
http://www.eventhelix.com/lte/presentations/3GPP-LTE-RLC.pdf
[32]
Evolved Universal Terrestrial Radio Access (E-UTRA); Packet Data
Convergence Protocol (PDCP) specification, 3GPP TS 36.323, 2011.
[33]
Evolved Universal Terrestrial Radio Access (E-UTRA); Radio Resource Control
(RRC); Protocol specification, 3GPP TS 36.331, 2012.
[34]
C. Han, “Security Analysis and Enhancements in LTE-Advance Networks”. Ph.D.
Dissertation, Dept. Mobile Systems Eng, Sungkyunkwan Univ, 2011.
[35]
F. Xiangning; C. Si; Z. Xiaodong; , "An Inter-Cell Interference Coordination
Technique Based on Users' Ratio and Multi-Level Frequency Allocations,"
International Conference on Wireless Communications, Networking and Mobile
Computing, pp.799-802, Sept. 2007.
[36]
Evolved Universal Terrestrial Radio Access (E-UTRA); Physical layer
procedures, 3GPP TS 36.213, 2011.
[37]
Roke, LTE MAC Scheduler & Radio Resource Scheduling [Online]. Available:
http://www.roke.co.uk/resources/white-papers/0486-LTE-Radio-Resource.pdf
[38]
Computerworld. 4G LTE networks hit battery life on some smartphones, Metrico
finds [Online]. Available:
http://www.computerworld.com/s/article/9226883/4G_LTE_networks_hit_battery
_life_on_some_smartphones_Metrico_finds
[39]
LTE; Evolved Universal Terrestrial Radio Access (E-UTRA); Radio Frequency
(RF) system scenarios, 3GPP TS 36.942, 2011.
[40]
D. Yu; W. Wen, "Non-access-stratum request attack in E-UTRAN," Computing,
Communications and Applications Conference (ComComAp), pp.48-53, Jan.
2012.
[41]
T. Hong, “Crosslayer Optimisation in LTE network to reduce the effect of
Co-channel Interference,” unpublished.
[42]
T. Ha, Theory and Design of Digital Communication Systems. Cambridge,
United Kingdom: Cambridge University Press, 2011.
77
THIS PAGE INTENTIONALLY LEFT BLANK
78
INITIAL DISTRIBUTION LIST
1.
Defense Technical Information Center
Ft. Belvoir, Virginia
2.
Dudley Knox Library
Naval Postgraduate School
Monterey, California
3.
Chairman, Code EC
Department of Electrical and Computer Engineering
Naval Postgraduate School
Monterey, California
4.
Tri T. Ha
Naval Postgraduate School
Monterey, California
5.
Weilian Su
Naval Postgraduate School
Monterey, California
6.
Tat Soon Yeo
Temasek Defence Systems Institute (TDSI)
National University of Singapore
Singapore
7.
Lai Poh Tan
Temasek Defence Systems Institute (TDSI)
National University of Singapore
Singapore
8.
Teo Tiat Leng
Defence Science and Technology Agency
Singapore
9.
Too Huseh Tien
Naval Postgraduate School
Monterey, California
79
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising