Local Area Network Dr. Farid Farahmand Revised on: 10/6/12

Local Area Network Dr. Farid Farahmand Revised on: 10/6/12
Local Area Network
Dr. Farid Farahmand
Revised on: 10/6/12
Data Network Areas
o 
o 
o 
WAN (Wide Area Networks)
MAN (Metropolitan Area Networks)
LAN (Local Area Networks)
n 
Sharing resources in small but geographically dispersed network
LAN Applications
o 
personal computer LANs
n 
n 
n 
o 
Potential issues for a single LAN
o 
o 
reliability
capacity
cost
High-speed office networks
o 
o 
high data rate
high speed interface
distributed access
limited distance
limited number of devices
backbone LANs
n 
n 
interconnect low-speed LANs
Resolve typical drawbacks to LANs
used particularly for desktop image
processing
o 
n 
o 
o 
o 
n 
interconnecting large systems
(mainframes and large storage devices)
printers, hard drives, etc.
o 
o 
backend networks
n 
low cost
limited data rate
share resources
o 
n 
o 
a single page with 200 pictures
elements (black and white) is about
3 Mbits!
high capacity local storage
LAN
Backbone
LAN
LAN Applications
o 
storage area networks (SANs)
n 
separate network to handle storage needs
o 
n 
n 
detaches storage tasks from specific servers
shared storage facility
o 
n 
n 
eg. hard disks, tape libraries, CD arrays
accessed using a high-speed network
o 
n 
à shared storage
eg. Fibre Channel
improved client-server storage access
direct storage-to-storage communication for backup
Project: Build a SAN!
Storage Area Networks
LAN Topologies
o 
Mesh Topology
n 
n 
o 
Star Topology
n 
n 
o 
n 
n 
All devices are connected to a central cable, called the bus
or backbone.
The bus is often terminated on both ends if not connected
to any devices.
The bus is typically duplex.
Ring Topology
n 
n 
o 
All devices are connected to a central switch/hub/repeater.
Nodes communicate across the network by passing data
through the switch/hub
Typically has unidirectional links
Bus Topology
n 
o 
Devices are connected with many redundant
interconnections between network nodes.
In a full mesh topology every node has a connection to
every other node in the network.
Mesh
All devices are connected to one another in the shape of a
closed loop, so that each device is connected directly to
two other devices, one on either side of it.
Closed loop with unidirectional links (links are point-topoint)
Tree Topology
n 
A hybrid topology. Groups of star-configured networks
are connected to a linear bus backbone.
http://www.webopedia.com/quick_ref/topologies.asp
Tree
Frame Transmission
Frame is absorbed
Frame is removed when it returns to
Its source
Ethernet – General
o 
o 
o 
o 
Most common LAN technology allowing multiple
devices to connect to each other and share resources
Developed by Xerox in 1970
Also known as IEEE 802.3
n  IEEE 802.3 Energy Efficient Ethernet Study Group
Each standards organization focuses on particular layers
of the protocol stack
n 
n 
n 
Institute for Electrical and Electronic Engineers (IEEE)
World Wide Web Consortium (W3C)
Internet Engineering Task Force (IETF)
Various Standard Emphasis
- Institute for Electrical and Electronic Engineers (IEEE)
- World Wide Web Consortium (W3C)
- Internet Engineering Task Force (IETF)
Other Standardization Bodies
o 
o 
o 
o 
Institute of Electrical and Electronics
Engineers (IEEE)
The European Computer Manufacturers
Association (ECMA)
The International Electrotechnical
Commission (IEC)
The International Organization for
Standardization (ISO).
IEEE 802 Model and Standards
o 
IEEE divides Layer 2 of the protocol stack into two
conceptual sub-layers
n 
The Logical Link Control (LLC)
o 
n 
sublayer specifies addressing and the use of addresses for
demultiplexing as described later in the chapter
The Media Access Control (MAC)
o 
sublayer specifies how multiple computers share underlying medium
IEEE 802 Model
and Standards
o 
IEEE assigns a multi-part
identifier of the form
XXX.YYY.ZZZ
n 
n 
n 
XXX denotes the category
of the standard
YYY denotes a
subcategory
If a subcategory is large
enough, a third level can be
added
Ethernet Protocol Architecture
802.2
802.3
EEE 802.2 is the IEEE 802 standard defining Logical Link Control (LLC)
EEE 802.3 is a collection of IEEE standards defining the Physical
Layer and Data Link Layer's media access control (MAC) sublayer of
wired Ethernet.
Ethernet Protocol Stack
Interface
Interface
EEE 802.2 is the IEEE 802 standard defining Logical Link Control (LLC)
EEE 802.3 is a collection of IEEE standards defining the Physical
Layer and Data Link Layer's media access control (MAC) sublayer of
wired Ethernet.
Ethernet Protocol Architecture
IEEE 802 Layers
o 
Physical
n 
n 
n 
n 
encoding/decoding of signals
preamble generation/removal
bit transmission/reception
transmission medium and
topology
IEEE MAC SUB-LAYER
Chapter 14
Multi-Access Protocols & Channel Allocations
o 
o 
o 
LAN technologies allow multiple computers to share
medium
n  any computer on the LAN can communicate with
any other
n  in order to share the medium we have to get
access
We use the term multi-access to describe the way
medium access is achieved
Thus, LAN is considered to be a multi-access
network
Channel (medium) Access Control Protocols
o 
o 
How are multiple
computers coordinated to
control (share) a single
medium?
In other words:
n 
What are the multi-access
protocols in multi-access
networks?
Permission is required
Just send it!
Take turn
Channelized Access Protocols
o 
Channelization refers to a mapping
between a given communication and
a channel in the underlying system
n 
There should be a mapping between
entities and a channel is referred to as
1-to-1 and static
o 
n 
Static channel allocation works well for
situations where the set of communicating
entities is known in advance and does not
change
A dynamic channel allocation scheme
can be established when a new station
appears, and the mapping can be
removed when the station disappears
Three main types of channelization
Also referred as
multiplexing techniques
Controlled Access Protocols
o 
Controlled access protocols provide a distributed
version of statistical multiplexing
n 
Controlled Access ProtocolsPolling
o 
o 
o 
Polling uses a centralized controller cycling through stations on the
network and gives each an opportunity to transmit a packet
The selection step is significant because it means a controller can choose
which station to poll at a given time
There are two general polling policies (how to select):
n 
Round robin order
o 
n 
Round-robin means each station has an equal opportunity to transmit packets
Priority order
o 
o 
Priority order means some stations will have more opportunity to send
For example, priority order might be used to assign an IP telephone higher
priority than a personal computer
Controlled Access ProtocolsReservation
o 
o 
It is often used with satellite transmission
Typically, reservation systems have a central controller that
follows Algorithm below
Reservation Algorithm
Controlled Access ProtocolsReservation
o 
o 
It employs a two-step process in which each round of packet
transmissions is planned in advance
In the first step
n 
o 
In the second step
n 
o 
each potential sender specifies whether they have a packet to send
during the next round, and the controller transmits a list of the
stations that will be transmitting
stations use the list to know when they should transmit
Variations exist
n 
where a controller uses an alternate channel to gather reservations for
the next round (out-of-band) - while the current round of
transmissions proceeds over the main channel
Controlled Access Protocols- Token Passing
o 
o 
It is most often associated with ring topologies
Although older LANs used token passing ring technology
n 
o 
Imagine a set of computers connected in a ring
n 
o 
the token circulates among all stations continuously
For a ring topology, the order of circulation is defined
n 
o 
and imagine that at any instant, exactly one of the computers has received a special control
message called a token
When no station has any packets to send
n 
o 
popularity has decreased, and few token passing networks remain
if messages are sent clockwise, the next station mentioned in the algorithm refers to the next
physical station in a clockwise order
When token passing is applied to other topologies (bus)
n 
n 
each station is assigned a position in a logical sequence
and the token is passed according to the assigned sequence
Random Access Protocols
o 
Some LANs do not employ a controlled access mechanism
n 
o 
o 
Instead, a set of computers attached to a shared medium attempt to access the
medium without coordination
The term random is used because access only occurs when a given station
has a packet to send
Three random access methods
n 
n 
n 
ALOHA
CSMA/CD (Collision Detection)
CSMA/CA (Collision Avoidance)
ALOHA
o 
An early network in Hawaii, known as
ALOHAnet, pioneered the concept of random
access
n 
o 
the network is no longer used, but the ideas have
been extended
The network consisted of a single powerful
transmitter in a central geographic location
n 
n 
It is surrounded by a set of stations/computer
Stations had a transmitter capable of reaching the
central transmitter
o 
o 
but not powerful enough to reach all the other
stations
ALOHAnet used two (2) carrier frequencies for
broadcasting:
n 
n 
one for outbound by the central transmitter to all
stations
and another for inbound by stations to the central
transmitter
Multiple outlying
stations
Outbound: 413.475 MHz
Inbound: 407.305 MHz
ALOHA
o 
The ALOHA protocol is straightforward:
n 
n 
o 
when a station has a packet to send it transmits the packet on the inbound
frequency
the central transmitter repeats the transmission on the outbound frequency
(which all stations can receive)
To insure that transmission is successful
n 
a sending station listens to the outbound channel
o 
o 
o 
Why might a packet fail to arrive? Interference
n 
if two stations simultaneously transmit (on the same frequency)
o 
o 
o 
if a copy of its packet arrives, the sending station moves to the next packet
if no copy arrives, the sending station waits a short time and tries again
the signals will interfere and the two transmissions will be garbled
called a collision, and say that the two transmitted packets collide
The protocol handles a collision
n 
by requiring a sender to retransmit each lost packet
ALOHA
Successful
Transmission
Packet collision
Retransmission is required
Three basic features:
Carrier Sense
Collision Detection
Binary Exp. Backoff
CSMA
o 
Ethernet requires each station to monitor the cable to detect whether another
transmission is already in progress
n 
n 
n 
o 
First listen for clear medium (carrier sense)
n 
n 
n 
n 
o 
If medium idle à transmit
If two stations start at the same instant à collision
Wait for reasonable time
If no activity then retransmit
Max utilization depends on propagation time (medium length) and
frame length
n 
o 
this process is known as carrier sense medium access
it prevents the most obvious collision problems
and substantially improves network utilization
Delay Types:
•  Node
•  Prop.
•  Trans.
Longer frame and shorter propagation gives better utilization
What if the line is busy?
n 
n 
n 
Non-persistent
1-persistent
P-persistent
Typically, propagation time is much less than transmission time
Non-persistent CSMA
Basic Idea:
1. 
If medium is idle, transmit; otherwise, go to 2
2. 
If medium is busy, wait amount of time drawn from probability
distribution (retransmission delay) and repeat 1
o 
Advantage: Random delays reduces probability of collisions
n 
Consider two stations become ready to transmit at same time while
another transmission is in progress:
o 
o 
If both stations delay same time before retrying, both will attempt to transmit
at same time
Disadvantage: Capacity is wasted because medium will remain idle
following end of transmission
n 
Even if one or more stations waiting – no one is transmitting
1-persistent CSMA
o 
o 
To avoid idle channel time, 1-persistent protocol is used
Station wishing to transmit listens and carries out the following:
1. 
2. 
o 
1-persistent stations are selfish
n 
o 
If medium idle, transmit; otherwise, go to step 2
If medium busy, listen until idle; then transmit immediately
If two or more stations waiting, collision guaranteed
Wasted time is shortened if frame length are long compared to
the propagation time
P-persistent CSMA
o 
Rules:
n 
If medium idle, transmit with probability p, and delay one time unit with
probability (1 – p)
n 
Time unit is typically equal to the maximum propagation delay
1. 
If medium busy, listen until idle and repeat step 1
If transmission is delayed one time unit, repeat step 1
2. 
What is an effective value of p?
Value of p?
o 
o 
o 
Avoid instability under heavy load
Consider that n stations are waiting to send, while a transmission is taking place
End of transmission:
n 
o 
Expected number of stations attempting to transmit = number of stations ready *
probability of transmitting = n*p
If np > 1 then on average there will be a collision
n 
n 
n 
Repeated attempts to transmit almost guaranteeing more collisions
Retries compete with new transmissions from other stations
Eventually, all stations trying to send
o 
o 
Continuous collisions; zero throughput
If np < 1 then
n 
n 
n 
If heavy load expected, p must be small
However, as p made smaller, stations wait longer
At low loads, this gives very long delays
CSMA/CD – Collision Detection
o 
o 
1. 
2. 
3. 
4. 
With CSMA, collision can occur for the entire duration
of transmission
Using CD stations listen whilst transmitting
If medium idle, transmit, otherwise, step 2
If busy, listen for idle, then transmit immediately
If collision detected, jam (transmit brief jamming signal)
then stop transmission – higher level of energy is sensed
After jam, wait random time (called backoff) then start
from step 1
Three basic features:
Carrier Sense
Collision Detection
Binary Exp. Backoff
Which Persistence Algorithm?
o 
o 
o 
IEEE 802.3 uses 1-persistent
Both non-persistent and p-persistent have performance
problems
Issues with 1-persistent (p = 1)
n 
Seems more unstable than p-persistent
o 
n 
n 
n 
Greed of the stations
But wasted time due to collisions is short (if frames long relative
to propagation delay)
With random backoff, unlikely to collide on next tries
To ensure backoff maintains stability, IEEE 802.3 and Ethernet
use binary exponential backoff
Binary Exponential Backoff
o 
A technique used by IEEE 802.3 and Ethernet to ensure that backoff
maintains stability
n 
o 
How does it work
n 
n 
n 
n 
o 
a computer chooses a random delay between 0 - d after one collision
a random delay between 0 - 2d after a second collision
a random delay between 0 - 4d after a third, and so on
After 16 unsuccessful attempts, station gives up and reports error
1-persistent algorithm with binary exponential backoff is efficient
over wide range of loads
n 
n 
o 
As congestion increases, stations back off by larger amounts to reduce
the probability of collision.
Low loads, 1-persistence guarantees station can seize channel once idle
High loads, at least as stable as other techniques
Backoff algorithm gives last-in, first-out effect
n 
Stations with few collisions transmit first
CSMA/CD Algorithm
CSMA/CD Algorithm
9.6 usec for 10MMbps
Collision Detection
o 
On baseband bus, collision produces much higher signal
voltage than signal
n 
n 
n 
o 
Collision detected if cable signal greater than single station signal
Signal attenuated over distance
Limit distance to 500m (10Base5) or 200m (10Base2)
For star-topology activity on more than one port is
collision
CSMA with Collision Avoidance
o 
CSMA/CD does not work as well in wireless LANs
n 
o 
A receiver that is more than a few hops away from
the transmitter
n 
o 
will not receive a signal, and will not be able to detect a
carrier
Wireless LANs use a modified access protocol
n 
o 
because a transmitter used in a wireless LAN has a limited
range
known as CSMA with Collision Avoidance (CSMA/CA)
The CSMA/CA triggers a brief transmission from
the intended receiver before transmitting a packet
CSMA with Collision Avoidance – Example (1)
n 
n 
n 
n 
n 
computer3 sends a short message to announce that it is ready to
transmit a packet to computer 2
and computer 2 responds by sending a short message announcing that
it is ready to receive the packet
all computers in range of computer 3 receive the initial announcement
and all computers in the range of computer 2 receive the response
as a result, even though it cannot receive the signal or sense a carrier,
computer 1 knows that a packet transmission is taking place
Hidden Station
Problem
CSMA with Collision Avoidance – Example (2)
o 
o 
Collisions of control messages can occur when using CSMA/CA, but
they can be handled easily
For example, if computer 1 and computer 3 each attempt to transmit a
packet to computer 2 at exactly the same time
n 
n 
o 
their control messages will collide
When a collision occurs, the sending stations apply random backoff before
resending the control messages.
Because control messages are much shorter than a packet, the probability
of a second collision is low
Signal propagation < span distance
Ethernet
o 
o 
IEEE 802.3 uses CSMA with 1-persistent
To ensure backoff maintains stability, IEEE
802.3 and Ethernet use binary exponential
backoff
Three basic features:
Carrier Sense
Collision Detection
Binary Exp. Backoff
Ethernet Addressing
Ethernet Addressing
o 
o 
IEEE has created a standard for addressing
Each packet that travels across the shared medium is intended for a
specific recipient
n 
o 
o 
The identifier is known as an address
Each computer is assigned a unique address
n 
o 
and each packet contains the address of the intended recipient
In the IEEE addressing scheme, each address consists of 48 bits; IEEE
uses the term Media Access Control address (or simply MAC address)
n 
o 
and only the intended recipient should process the packet
networking professionals often use the term Ethernet address
IEEE allocates a unique address for each piece of interface
n 
Each Network Interface Card (NIC) contains a unique IEEE address assigned
when the device was manufactured
Ethernet Addressing
o 
Remember: MAC address is a 48 bit flat addressing
n 
o 
When a device is added its MAC address is announced to others
n 
o 
It is not hierarchical as in IP addressing
For same LLC, may have several MAC options
Logical Link Control
n 
n 
Interface to higher levels
Flow and error control
Data Link Layer
LLC (Logical Link) Layer
MAC (Media Access) Layer
Network Layer
Physical Layer
Ethernet Addressing
o 
MAC address is 48 bits:
n 
n 
24 bits (OUI – Organizationally unique Identifier
24 bit hardware address – burned in the ROM
Ethernet Addressing
My OUI
Ethernet Addressing
o 
o 
o 
The IEEE addressing supports three types of addresses that correspond to three
types of packet delivery
n 
Unicast, multicast, broadcast
The standard specifies that a broadcast address consists of 48 bits that are all 1s
n 
Thus, a broadcast address has the multicast bit set
Broadcast can be viewed as a special form of multicast
n 
Each multicast address corresponds to a group of computers
n 
Broadcast address corresponds to a group that includes all computers on the
network
Packet Processing and Efficient MultiPoint Delivery
o 
o 
Recall that a LAN
transmits packets over a
shared medium
In a typical LAN
n 
n 
n 
n 
each computer on the
LAN monitors the shared
medium
extracts a copy of each
packet
and then examines the
address in the packet
determine whether the
packet should be
processed or ignored
Packet Processing Algorithm in a LAN
Frames and Framing
o 
o 
o 
Frame
Metadata
Payload
Framing refers to the structure added to a sequence
of bits or bytes that allows a sender and receiver to
agree on the exact format of the message
Ethernet is a packet-switched network that transmits
and receives Ethernet frames - each frame
corresponds to a packet
A frame consists of two conceptual parts:
n 
Header that contains metadata, such as an address
o 
n 
contains information used to process the frame
Payload that contains the data being sent
o 
o 
contains the message being sent
and is usually much larger than the frame header
Frames and Framing
o 
A message is opaque
n 
n 
o 
Opaque V.s Forwarding
n 
o 
in the sense that the network only examines the frame header
the payload can contain an arbitrary sequence of bytes that are only
meaningful to the sender and receiver
Forwarding checks the CRC; Opaque just checks the destination
Some technologies delineate each frame by sending a short
prelude before the frame and a short postlude after it
Frames and Framing
o 
Assume that a packet header consists of 6 bytes
n 
o 
the payload consists of an arbitrary number of bytes
We can use ASCII character set
n 
n 
the Start Of Header (SOH) character marks the beginning of a frame
and the End Of Transmission (EOT) character marks the end
Frames and Framing
o 
o 
o 
o 
In the ASCII character set
n  SOH has hexadecimal value 201
n  EOT has the hexadecimal value 204
An important question arises
n  what happens if the payload of a frame includes one or more bytes
with value 201 or 204?
The answer lies in a technique known as byte stuffing
n  that allows transmission of arbitrary data without confusion
Examples of bit stuffing:
n  the sender replaces each occurrence of SOH by the two characters
ESC [1B hex] + A
n  each occurrence of EOT by the characters ESC + B
n  and each occurrence of ESC by the two characters ESC + C
http://www.asciitable.com/
Ethernet Frame Format
Max. Length: 8+6+6+2+1500+4=1526
Min Length: 8+6+6+2+46+4=72
8 Bytes of Preamble
Nothing is being Sent!
Ethernet Frame Format
Calculations
8 Bytes of Preamble
Nothing is being Sent!
Max. Length: 6+6+2+1500+4=1518 byte excluding the preamble
Min Length: 6+6+2+46+4=64 byte excluding the preamble
When transmitting consecutive Min. Length frames, actually each frame occupies 72+12 = 84 Bytes
Assuming 10 Mbps Ethernet (RJ-45 Cat 5 UTP) à Inter-frame gap will be 9.6 usec
Assuming we have a 100 Mbps Ethernet connection sending minimum size frames,
Overhead Ratio: (Total Bytes Sent – Data Byte) / Total Bytes Sent = (84 B – 64 B)/84 B = 23.8 %
thus, assuming no other traffic and no collision, total BW Efficiency will be only 23.8 %.
But typical Ethernet packet size is about 260 Bytes (lots of TCP ACK packets!)
Framing Structure
Exercise: Look at Frame # 9 and Frame 4: Identify all the fields in the frame.
ARP Packet
Note that the type is ARP
Given MACà what is IP
Ethernet Frame Format
o 
o 
o 
o 
o 
o 
o 
o 
Preamble (PRE)- 7 bytes. The PRE is an alternating pattern of ones and zeros that tells
receiving stations that a frame is coming, and that provides a means to synchronize the
frame-reception portions of receiving physical layers with the incoming bit stream.
Start-of-frame delimiter (SFD)- 1 byte. The SOF is an alternating pattern of ones and
zeros, ending with two consecutive 1-bits indicating that the next bit is the left-most bit in
the left-most byte of the destination address.
Destination address (DA)- 6 bytes. The DA field identifies which station(s) should receive
the frame..
Source addresses (SA)- 6 bytes. The SA field identifies the sending station.
Length/Type- 2 bytes. This field indicates either the number of MAC-client data bytes that
are contained in the data field of the frame, or the frame type ID if the frame is assembled
using an optional format.
Data- Is a sequence of n bytes (46=< n =<1500) of any value. (The total frame minimum is
64bytes.)
Frame check sequence (FCS)- 4 bytes. This sequence contains a 32-bit cyclic redundancy
check (CRC) value, which is created by the sending MAC and is recalculated by the
receiving MAC to check for damaged frames.
Inter-Frame Gap: Used to ensure that a single node does not utilize the link at all the time.
Ethernet Type Field
o 
The type field in an Ethernet frame provides
multiplexing and demultiplexing (2 bytes
long)
n 
o 
The protocols used on the Internet send IP
datagrams and ARP (Address Resolution
Protocol) messages over Ethernet
n 
n 
o 
Allows a given computer to have multiple
protocols operating simultaneously
Each is assigned a unique Ethernet type
(hexadecimal 0800 for IP datagrams and
hexadecimal 0806 for ARP messages)
When transmitting a IP datagram in an
Ethernet frame, the sender assigns a type 0800
When a frame arrives at its destination
n 
the receiver examines the type field, and it uses
the value to determine which software module
should process the frame
Ethernet Types:
Versions of Ethernet
o 
o 
IEEE developed a standard for Ethernet (1983) and attempted
to redefine the Ethernet frame format
The IEEE working group that produced the standard is
numbered 802.3
n 
o 
professionals often refer to it as 802.3 Ethernet
The major difference between conventional Ethernet and
802.3 Ethernet arises from the interpretation of the type field
n 
n 
The 802.3 standard interprets the original type field as a packet
length, and adds 8-byte header that contains the packet type
The extra header is known as a Logical Link Control / Sub-Network
Attachment Point (LLC/SNAP) header; (next slides)
http://standards.ieee.org/getieee802/download/802-2001.pdf
Ethernet Types
How to distinguish
o 
Novell's non-standard variation of IEEE 802.3
("raw 802.3 frame") without an IEEE 802.2
LLC header.
n 
n 
o 
The Ethernet Version 2 or Ethernet II frame,
the so-called DIX frame (named after DEC,
Intel, and Xerox)
n 
n 
o 
Uses Length Field + LLC
IEEE 802.2 LLC/SNAP frame
n 
n 
o 
often used directly by the Internet Protocol.
Uses Ethernet Type -Larger than 05DC Hex à
IEEE 802.2 LLC frame
n 
o 
If the IPX header (0xFF-FF) - the data field
Length Field is used (Max. 1500=05DC Hex)
Uses Length Field + LLC + SNAP
If SSAP value is 0xAA, the frame is interpreted as a
SNAP frame otherwise LLC only
Example of Ethernet Version 2:
http://en.wikipedia.org/wiki/Ethernet
Ethernet Types:
Comparing IEEE802.3
conventional Ethernet
(Ethernet Version II
or Ethernet II frame)
Max. 1518 byte
0800/0806 hex
IEEE 802.2 LLC/SNAP
Frame (Max. 1518 byte)
Added 8 bytes by 802.3
IEEE 802.3 Frame Format
LLC/SNAP
o 
Remember: the upper sublayer of the Data Link Layer is LLC
(layer 2)
n 
n 
n 
o 
Destination SAP
Source SAP
Control (Type I: Connectionless
Type II: Connection Oriented )
HDLC (High Level Link Control Protocol) is a general purpose data link
layer
HDLC uses the services of a physical layer:
It offers best effort or reliable communications path between the TX & RX
LLC provides multiplexing and flow control mechanisms
that make it possible for several network protocols (IP, IPX)
to coexist within a multipoint network (different SAP) and to
be transported over the same network media
n 
n 
n 
n 
LLC field is divided into DSAP, SSAP, & Control
Service Access Point (SAP): 8-bit 802.2 fields are typically used in data link
layers (LLC sublayer) for addressing purpose –
SAP is the label used for the equipment
E.g., ATP SONET/SDH have their own SAP
LLC/SNAP
o 
o 
o 
o 
There are different Service Access Points (SAPs): Transport SAP, Session
SAP, Network SAP (NSAP)
Note that NSAP in OSI model is similar (broadly speaking) to IP Address
in TCP/IP Layered Model
TSAP in OSI model serves similar task as TCP Port Address
The Subnetwork Access Protocol (SNAP) is a mechanism for
multiplexing, on networks using IEEE 802.2 LLC, more protocols than
can be distinguished by SAP
n 
n 
SNAP supports vendor-private protocol identifier spaces (OUI –
organizationally unique identifier) .
Identifies protocols by Ethernet type field values;
o 
IEEE 802.3, IEEE 802.4, IEEE 802.5, IEEE 802.11 and other IEEE 802 physical
network layers, as well as with non-IEEE 802 physical network layers such as
FDDI that use 802.2 LLC.
Ethernet Advantages and Disadvantages
o 
Advantages
n 
o 
Easy to setup; Requires no configurations; Robust to noise
Disadvantages
n 
High collision rate, hence it must operate under low load conditions
o 
o 
n 
Providing non-deterministic service
o 
o 
n 
Packets may experience indefinite delay due to high collision rate
Ethernet may not be suitable for applications demanding a bound on worst-case delay
Ethernet, and in general CSMA, does not support priority
o 
o 
n 
As the load increases the throughput decreases to zero (that is amount f traffic shifted
from one node to another in unit of time)
Ethernet loads rarely exceed 30%, so this is not a major problem
Each station has equal chance to transmit
In case of client-server we may actually want the server to have greater priority
Requiring minimum packet length of 64 Bytes
o 
This can increase the overhead on applications sending only 1-5 bytes!
Ethernet Evolution
Ethernet Evolution - Thicknet
o 
Ethernet has undergone several major changes
n 
o 
The original Ethernet wiring scheme was informally called thick wire
Ethernet or Thicknet
n 
n 
o 
with the most significant changes in media and wiring
because the medium consisted of a heavy coaxial cable
the formal term for the wiring is 10Base5
Hardware used with Thicknet was divided into two major parts
n 
n 
A NIC handled the digital aspects of communication
A separate electronic device called a transceiver connected to the Ethernet
cable
o 
o 
o 
It handles carrier detection, conversion of bits into appropriate voltages for
transmission, and conversion of incoming signals to bits
A physical cable known as an Attachment Unit Interface (AUI)
connected a transceiver to a NIC in a computer
A transceiver was usually remote from a computer
Ethernet Evolution - Thicknet
o 
Hardware used with Thicknet was divided into two major parts
n 
n 
A NIC handled the digital aspects of communication
A separate electronic device called a transceiver connected to the Ethernet
cable
o 
o 
o 
It handles carrier detection, conversion of bits into appropriate voltages for
transmission, and conversion of incoming signals to bits
A physical cable known as an Attachment Unit Interface (AUI)
connected a transceiver to a NIC in a computer
A transceiver was usually remote from a computer
ThickNet
Thinnet Ethernet Wiring
o 
A second generation of Ethernet used a thinner coaxial cable that was more
flexible than Thicknet
n 
o 
Formally named 10Base2 and informally known as Thinwire Ethernet or Thinnet
Thinnet integrates a transceiver directly on the NIC
n 
runs a coaxial cable from one computer to another
© 2009 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.
70
Twisted Pair Ethernet
Wiring and Hubs
o 
A third generation of Ethernet wiring made a dramatic shift:
n 
In place of coaxial cable
o 
n 
Instead of heavy, shielded cabling
o 
o 
it uses twisted pair wiring
The technology is informally known as twisted pair Ethernet, and has replaced other
versions
n 
o 
it uses a central electronic device separate from the computers attached to the network
Thus, when someone now refers to Ethernet, they are referring to twisted pair Ethernet
The electronic device that served as the central interconnection was known as a hub
n 
Hubs were available in a variety of sizes, with the cost proportional to size, but recently replaced
with switches
Logical and Physical Topologies
o 
To understand Ethernet topology
n 
n 
we must distinguish between logical and physical topologies
Logically
o 
n 
twisted pair Ethernet employs a bus topology
Physically
o 
twisted pair Ethernet forms a star-shaped topology
Twisted pair Ethernet Technologies
Compatibility
o 
Significant improvements have been made in the quality and
shielding available in twisted pair cables
n 
o 
o 
the data rate used on twisted pair Ethernet has increased (three types –
above)
Higher-speed Ethernet technologies use an electronic device
known as a switch rather than a hub
To remain backward compatible
n 
n 
n 
standards for the higher-speed versions specify that interfaces automatically
sense the speed at which a connection can operate
and slow down to accommodate older devices
For example, if one plugs an Ethernet cable between an old device that uses
10BaseT and a new device that uses 1000BaseT
o 
the new device will autosense
Ethernet Specifications
o 
Ethernet classification
< Speed> <Baseband/Broadband> <Physical>
o 
Speed: 3,10,100 Mbps depending on the cable type:
n 
n 
o 
Baseband is typically used for small building
Broadband is used for larger networks such as Cable TV
Cable Standards
YYBase-xx
Data rate: 3, 10,100
Baseband, carrying a single data
10BASE-T
http://www.comptechdoc.org/independent/networking/cert/netphysical.html
Cable type:
T = Twisted pair
2 = Coax
5 = Thick coaxial
F,X= Fiber optics
Physical Interfaces
o 
o 
Twisted Pair
n 
Unshielded Twisted Pair (UTP). Normally UTP
n 
Shielded twisted pair (STP) – e.g., Cat5E Shielded Twisted Pair
Coaxial
n 
n 
n 
n 
n 
n 
n 
n 
o 
RG-62 - 93 ohm, primarily used for ArcNet.
RG-59 - 75 ohm, for broadband transmission such as cable TV.
RG-58 /U - 50 ohm, with a solid copper wire core for thin Ethernet.
RG-58 A/U* - 50 ohm, with a stranded wire core.
RG-58 C/U* - Military version of RG-58 A/U.
RG-11 - 75 ohm thick Ethernet.
RG-8 - 50 ohm thick Ethernet.
RG-6 - Used for satellite cable (if you want to run a cable to a satellite!).
Fiber-optic
n 
expensive taps / better alternatives available / not used in bus LANs
n 
Single mode cables for use with lasers and offer greater bandwidth and
costs more
Multimode cables for use with Light Emitting Diode (LED) drivers
n 
"RG" was originally a unit indicator for bulk RF cable in the U.S. military
http://www.ciscopress.com/articles/article.asp?p=31276&seqNum=2
50 Ohms Cable - RG58/U--50 Ohm
AWG Standards
American wire gauge (AWG)
Silver
Smaller
Higher
(less signal goes through)
Ethernet Cables
o 
Ethernet Crossover cable
n 
n 
n 
o 
Used to connect computing devices (PCs, two routers, or two
hubs) together directly
Used to connect PC and router
Example: connecting two personal computers via their network
adapters
Ethernet Standard straight through cable
n 
n 
n 
Each pin of the connector on one end is connected to the
corresponding pin on the other connector.
used to connect a PC or Router or an Ethernet hub (connecting
different equipments)
Exception: Connecting an uplink port of a hub to a regular port
of another regular port of hub
Wiring Serial
Connection
o 
o 
o 
Asynchronous Serial Ports are used to access the
console port of a router a PC– used to send ASCII
characters
The serial port is connected to a PC via RS-232
(DB-25 connector with 25 pins or DB-9 connector
with 9 pins or RJ-45 with 8 pins – similar to the
Ethernet cable)
Rolled cable (rollover) is used to access the console
port
n 
Pin 1 à 8 and so on (all pins are rolled)
Example
o 
Which interface uses a
crossover cable?
n 
n 
n 
n 
1-?
2-?
3-?
4-?
Example
o 
Which interface uses a
crossover cable?
n 
n 
n 
n 
1-Rolled cable
2-Crossover cable
3-Straight-through cable
4-Straight-through cable
Fiber Optics Cable
Optical Connections
o 
o 
o 
Different physical contacts are used for optical
fiber cables
A typical optical fiber cable usually includes
several optical fibers around a central steel cable.
Various protective layers are applied, depending on
the harshness of the environment where the cable
will be situated.
Ethernet Cable Types (Examples)
o 
o 
o 
10BASE-2
n 
Also known as "Thin Ethernet" or Thinnet, 10BASE-2 is an
n 
Also known as "Thick Ethernet" or Thicknet, 10BASE-5 is an IEEE
standard for baseband Ethernet at 10MBps over thick coaxial cable.
10BASE-5 has a maximum distance of 500 meters.
IEEE standard for baseband Ethernet at 10MBps over thick
coaxial cable. 10Base2 has a maximum distance of 185 meters.
10BASE-5
10BASE-T
n 
o 
Similar to the standard telephone cabling, 10BASE-T is a 10MBps
CSMA/CD Ethernet LAN that works on Category 3 or better
twisted-pair cables capable of being 100 meters long.
100BASE-T / Fast Ethernet / 100BASE-TX (or 802.3u)
n 
Fast Ethernet is also referred to as 100BASE-T or 802.3u and is a
communications protocol that enables computers on a local-area
network to share information with one another at rates of 100
million bits per second instead of the standard 10 million BPS. Fast
Ethernet works over Category 5 twisted-pair wiring.
Cable type:
T = Twisted pair
2 = Coax
5 = Thick coaxial
F,X= Fiber optics
100BASE-TX is the predominant form of Fast Ethernet, and runs over two
wire-pairs inside a category 5 or above cable (a typical category 5 cable
contains 4 pairs and can therefore support two 100BASE-TX links)
Common Ethernet Speeds: 10BaseT Ethernet; Fast Ethernet; GigE
STP vs UTP
Cat 5 (UTP) Un-Shielded Twisted Pair
Shielded Twisted Pair
Fast Ethernet - IEEE 802.3u
o 
In 1995 a committee was established to develop
standards for faster LAN
n 
n 
o 
o 
o 
Backward compatible
10 nsec bit time (100 Mbit/sec)
Interesting properties
Three signal levels : +V, 0, -V
Codewords are selected such that line is d.c.balanced
è all codewords have a combined weight of 0 or 1.
Common Ethernet Speeds: 10BaseT Ethernet; Fast Ethernet; GigE
Gigabit Ethernet – 802.3z
o 
Started in 1998 following formation of a High-Speed Study
Group to study convening packets in Ethernet format at
Gbps speed
n 
n 
n 
o 
Used to interface lower-speed Ethernets (LANs)
Suitable for streaming HD and multimedia
Backward compatible
Compatible with Fast Ethernet
n 
n 
n 
Using similar CDMA/CD frame format and MAC protocol
Point-to-point only (not multipoint)
Supports Full duplex & half duplex
Good Summary: http://www.infocellar.com/networks/standards/ethernet.htm
10G Ethernet
o 
o 
Operates at 10G
Several IEEE standards, including:
n 
802.3ae-2002 (fiber -SR, -LR, -ER, etc.)
o 
n 
n 
SR: Short Reach; LR: Long Reach; ER: Extended
Reach; LX: Fiber;
802.3ak-2004 (-CX4 copper cable)
802.3an-2006 (10GBASE-T copper twisted pair),
2-Pair & 4-Pair CAT 5
o 
o 
o 
o 
Ethernet 10Base-T uses pairs 2 and 3 (pins 1-2, 3-6)
Ethernet 100Base-T4 uses pairs 2 and 3 (pins 1-2, 3-6) – using 4 wires
Ethernet 100Base-T8 uses pairs 1,2,3 and 4 (pins 4-5, 1-2, 3-6, 7-8) – using 8 wires
GigE 1000Base-TX uses pairs 1,2,3 and 4 (pins 4-5, 1-2, 3-6, 7-8) – using 8 wires
Differential TX & RX
Differential TX & RX
Including +/- 15V and GND
RJ-45 Pinout
Physical Media Comparisons
Cable
Category
Speed
Notes
Media
Distance
(meters)
Speed
1
None
Used for old telephone
systems
UTP
100
4-100Mbps
2
4Mps
STP
100
16-155Mbps
3
10Mps
Thinnet
185
10Mbps
4
16Mps
Thicknet
500
10Mbps
5
100Mps
Fiber
2000
100Mbps-2Gbps
The minimum category for
data networks
Cat 5 network cable, used by
most networks today
6
Data patch, Two pair with foil
and braided shield
7
Undefined
8
Flat cable for under carpets
with two twisted pair
9
Plenum cable with two
twisted pair. It is safe if you're
having a fire.
The Electronic Industries Association and
Telecommunications Industries Association (EIA/TIA)
defined a standard called EIA/TIA 568 which is a
commercial building wiring standard. It defines
transmission speed and twists per foot.
Gbit Ethernet Medium Options (log scale)
100BASE-T Options
100BASE-TX: two pairs of high-quality twisted-pair wires
100BASE-T4: four pairs of normal-quality twisted-pair wires
100BASE-FX: fiber optic cables
Example
High-Speed
10/100Mbps
Workgroup Switch
The PLANET FSD-1605 /
FSD-2405 is a
10/100Mbps Fast
Ethernet Switch with
16 / 24 ports
A
Questions:
What rate can 1000Base-T handle?
What type of an Ethernet Switch is the Backbone Switch?
What type of cable will you use to connect to clients?
What type of cable would you use to connect to the servers?
How many total number of clients can be supported?
Show the route for a packet entering the router with destination A
Assuming no more FSDs can be added, can you support 100 clients? Explain!
Example
High-Speed
10/100Mbps
Workgroup Switch
The PLANET FSD-1605 /
FSD-2405 is a
10/100Mbps Fast
Ethernet Switch with
16 / 24 ports
Layer 3 Switch
A
Questions:
What rate can 1000Base-T handle? – 1 Gig bits per sec
What type of an Ethernet Switch is the Backbone Switch? – Fast Ethernet
What type of cable will you use to connect to clients? – Two Cat 5 UTP or 2 STP (100Base-TX)
What type of cable would you use to connect to the servers? – 2 Cat 5 / UTP
How many total number of clients can be supported? - 64
Show the route for a packet entering the router with destination A – shown above
Assuming no more FSDs can be added, can you support 100 clients? Explain! à Fairness and
system utilizations must be considered as we add more new hubs to support more clients.
Another Example
A
Backbone Example
Another Example
Backbone Example
A
Layer 2 hub
Devices
Data Communication Equipments
n 
n 
n 
n 
Used to transport information
over the network
Perform data formatting, routing,
transporting, and ensure to
maintain data integrity
Examples: Routers, Switches,
Bridges
Differ in terms of cost, size,
carries class (reliability or
uptime requirement)
A generic Point-to-point system
DCE
DTE
DTE
DTE (Data Terminating Equipment)
-e.g., PC, Mainframe,
where signal ends
DCE (Data Communications Equipment)
-T1, POTS, ISDN
Switching Equipment
o 
Network Switching
n 
n 
o 
Different/same networks
Example: Routers, Gateways, etc.
LAN Switching
n 
n 
Different/same Segments (collision domains)
Layer 2 switch, Repeater), Hub, etc.
A Different View:
LAN Switching Technologies
o 
o 
o 
o 
LAN Switching involves packet switching in LAN
Different technologies
n  Speed, Addressing, Utilization, etc.
Layer 1 switching – just blindly copying the packet
n  Pass-through devices
n  Considered to be analog
n  Example is a hub or repeater
Layer 2 switching is hardware-based switching
n  Uses the media access control address (MAC address) from the
host's network interface cards (NICs) to decide where to forward
n  Allows dividing the LAN into multiple Segments
n  Typically uses 80/20 rule (80 percent of the traffic is local)
n  Not all bridges support multicasting and broadcasting
n  Example: Multiport Bridge (Layer 2 Switch)
Interconnecting LANs
o 
In many scenarios we need larger LANs
n 
n 
o 
Connecting similar LANs
Creating longer LANs
Interconnecting using repeaters, bridges, or
routers
Layer 1 LAN Devices
o 
Repeaters
n 
n 
o 
Boosts the signal (layer 1 operation)
Transparent to the signal
10BaseT
(185 Meter)
Hub
n  active central element of star layout
n  each station connected to hub by two-pair
UTP lines
n  hub acts as a repeater (layer 1 operation)
n  limited to about 100 m by UTP properties
n  optical fiber may be used for longer
n  physically star, logically bus
n  transmission from a station seen by all
others
n  if two stations transmit at the same time
have a collision
Repeater
10BaseT
(185 Meter)
Hub
Router
Repeater
Hub
Router
Collision!
Hub and Repeaters
Single Collision/Broadcast Domain
HUB
Hub Stack Configuration
Reptr
HUB
Hub and Repeaters
HUB
Hub Stack Configuration
Reptr
HUB
Cisco's Catalyst Switch
Layer 2 Switch
o 
An Ethernet switch, sometimes called a
Layer 2 switch is an electronic device
that resembles a hub
n 
n 
o 
a switch provides multiple ports that each
attach to a single computer
and a switch allows computers to send
frames to one another
The difference between a hub and a
switch arises from the way the devices
operate:
n 
n 
n 
n 
a hub operates as an analog device that
forwards signals among computers
while a switch is a digital device that
forwards packets
We can think of a hub as simulating a
shared transmission medium
We think of a switch as simulating a
bridged network that has one computer
per LAN segment
As before
Layer 2
Switch
Capacity
o 
HUB uses star wiring to
attach stations
n 
n 
n 
transmission from any
station received by hub
and retransmitted on all
outgoing lines
only one station can
transmit at a time
total capacity of LAN is 10
Mbps
Compare the traffic capacity
of a hub and a switch
Capacity of Layer 2
Switch
o 
HUB can improve performance using a layer 2
switch
n 
n 
n 
o 
Layer 2 switches can convert bus LAN or hub LAN
to switched LAN
n 
o 
e.g. Ethernet LANs use Ethernet MAC protocol
Layer 2 switches have dedicated capacity equal to
original LAN
n 
o 
can switch multiple frames between separate ports
multiplying capacity of LAN
Hardware-based addressing
assuming switch has sufficient capacity to keep up
with all devices
Layer 2 switches scale easily
n 
additional devices attached to switch by increasing
capacity of layer 2
Interconnection Between Different
Networks
Layer 1, 2, and 3
Layer
3
Router
Routers
o 
o 
o 
o 
o 
o 
o 
o 
o 
o 
o 
Handles layer 2 and 3 operations (complex)
Cracks packets to check their destinations
Performs load balancing
Detects failure and performs dynamic routing
Provided traffic statistics
Support different interfaces: OC, 56 Kbps, Ethernet, DS3, etc.
Exchanges routing information
Routers have their own address
Require lots of memory
Uses discovery protocols to find all neighboring nodes
Types
n 
n 
Intermediate routers: Connecting two LANS - Translating packet from
one LAN to another (Ethernet Frame to Token Ring Frame)
Gateway Routers: Connecting a LAN to the Internet
Bridging Devices
o 
Bridging allows connecting
identical physical / link layer
protocols
n 
n 
n 
minimal processing
can map between MAC formats
reasons for use
o 
o 
o 
o 
Reliability – partitioning the
network
Performance – smaller LANs
perform better
Security – supporting different
traffic types by each LAN
Geography – separated
o 
Bridge operation
n 
n 
n 
n 
n 
n 
o 
o 
n 
In order to handle layer 2 routing,
bridging must include ways to discover
which devices are connected together
& and how to forward
Used to connect two/more
LANS
Less sophisticated than routers
Separates different segments
Its routing protocol architecture
is based on 802.1D
More intelligent than repeaters
(operates digitally)
Connect layer 2 segments
Handles layer 2
functionalities only
Transparent to layer 3
Uses polling and discovery
packets
o 
o 
What is your MAC address
again?
Establishes a tree-system
Conceptual use of bridges in a switch
A switch consists of an intelligent interface attached to each port
and a central fabric that provides simultaneous transfers
An interface contains a processor, memory, and other hardware needed to
accept a packet consult a forwarding table and send the packet across the fabric
to the correct output port
An interface can buffer arriving packets when an output port is busy
Bridge Operation
Hub (segment 1)
Bridge
The transmission can be within segment 1 only
Or flooded between segment 1 and 2
Hub (segment 2)
Bridge Operation
Checks the MAC address (No LLC)
Hub (segment 1)
Bridge
The transmission can be within segment 1 only
Or flooded between segment 1 and 2
Remember: bridge uses MAC addresses to
perform filtering – layer 2 switch
Hub (segment 2)
Bridge Design
o 
the bridge listens in promiscuous mode on each segment
n 
o 
o 
o 
o 
o 
o 
o 
i.e., receives all packets sent on the segment
no modification to frame content or format
no encapsulation
exact bitwise copy of frame
minimal buffering to meet peak demand
contains routing and address intelligence
may connect more than two LANs
bridging is transparent to stations (Cut-through)
Bridges and Frame Filtering
o 
Bridges do not blindly forward a copy of each frame
from one LAN to another
n 
o 
A bridge examines the destination address in a frame
n 
o 
Instead, a bridge uses MAC addresses to perform filtering
– layer 2 switch
and does not forward the frame onto the other LAN
segment unless necessary
If the LAN supports broadcast or multicast
n 
the bridge must forward a copy of each broadcast or
multicast frame
o 
to make the bridged LAN operate like a single LAN
About Routing Protocols
o 
How can a bridge know which computers are
attached to which segments?
n 
Most bridges are called adaptive or learning
bridges
o 
n 
because they learn the locations of computers
automatically
To do so, a bridge uses source addresses
Here is how….
Bridge Protocol Architecture
o 
o 
o 
IEEE 802.1D
MAC level
Bridge does not need LLC layer
Bridges and
LANs with
Alternative
Routes
Bridge Protocol Architecture
o 
Bridges must have some routing capacity
n 
n 
n 
o 
Must know the topology
Capable of changing the routing when changes occur
Route based on the MAC address
Routing
n 
n 
n 
Fixed routing (for small and stable networks)
Spanning Tree (802.1)
Source routing (802.5 – Token Ring)
Remember: 802.1D is the IEEE MAC
Bridges standard which includes Bridging,
Spanning Tree and others.
Fixed Routing
o 
Used for each source-destination pair of LANs
n 
n 
n 
n 
done in configuration
usually least hop route
only changed when topology changes
widely used but limited flexibility
Example of Fixed Routing
Create a routing matrix – stored in the bridge
Shortest route is based on least cost function
Note that AàB = BàA
o 
o 
o 
A
A
B
NONE
B
101
C
102
D
E
F
o 
C
101
D
F
102
NONE
103
104
NONE
103
107
E
105
NONE
104
NONE
105
NONE
Find routing from Eà F
The table is manually created for each bridge!
Example of Fixed Routing
Create a routing matrix – stored in
the bridge
o 
A
A
B
NONE
B
101
C
102
D
E
F
o 
o 
o 
C
101
D
F
102
NONE
103
104
NONE
103
107
E
105
NONE
104
NONE
105
NONE
Find routing from Eà F
EàBàAàCàF
EàAàCàF (lower hop count)
The table is manually created for each bridge!
Use Shortest
Path Routing
Spanning Tree
o 
o 
o 
bridge automatically develops routing table
automatically updates routing table in response to
changes
three mechanisms:
n 
n 
n 
frame forwarding
address learning
loop resolution (distributed spanning tree)
Frame ForwardingForward a request or block?
o 
When a frame arrives, the bridge must
extract the MAC address from the frame
n 
o 
The bridge must maintain forwarding
database for each port
n 
o 
use the address to determine whether to forward
the frame
n 
n 
n 
Broad N
cast
Found
MAC?
lists station addresses reached through each port
For a frame arriving on port X:
n 
Check MAC
Address
search forwarding database to see if MAC
address is listed for any port except X
if address not found, forward to all ports except
X
if address listed for port Y, check port Y for
blocking or forwarding state
if not blocked, transmit frame through port Y
Blocked?
N
Forward to Y
Address Learning –
Who is connected to the bridge
o 
o 
A bridge learns that a computer is present on a
segment as soon as the computer transmits a frame
When a frame arrives from a given segment
n 
the bridge extracts the source address from the header and
adds the address to a list of computers attached to the
segment
Address
Learning
(No looping)
Port 2
Port 1
Only src addresses are detected
Initially, the
Bridge looks at
both Segments to
forward the
frame
At this point the bridge
Knows all the connected nodes to its two segments
Address
Learning
(No looping)
Port 2
Port 1
A tree with no loop!
B
A graph
representation of the topology!
H1
A
B
H2
C
X
Y
Z
Distributed Spanning Tree Algorithm
(Loop Resolution)
o 
o 
o 
o 
address learning works for tree layout
however, in general graph have loops
for any connected graph there is a
spanning tree maintaining connectivity
with no closed loops
IEEE 802.1 Spanning Tree Algorithm
Graph w /
Loops
Spanning
Tree
No Closed
Loop
- Address learning does not work
when there is a loop!
-  In the case above, bridges get
confused!
-  Both bridges want to
send packet from A to B!
Spanning Tree Protocol
o 
STP consists of three steps:
1. 
Root election
o 
o 
2. 
Shortest path computation
o 
o 
3. 
Each bridge computes a shortest path to the root bridge.
Links included in the shortest paths of all bridges form the spanning tree
Frame Forwarding
o 
o 
To permit a manager to control the election a bridge ID is used; it
consists of two parts: a 16-bit configurable priority number and a 48-bit
MAC address
bridges multicast a packet that contains their bridge ID, and the bridge
with the smallest ID is chosen ( sometimes only priority number is
picked) See http://www.cisco.com/warp/public/473/spanning_tree1.swf
An interface that connects to the shortest path is enabled for forwarding
packets; an interface that does not lie on the shortest path is blocked,
In STP, Ethernet bridges communicate amongst themselves
using a multicast address that is reserved for STP
Example:
Resolving the Loop of Bridges
- To prevent cycles, a Distributed Spanning
Tree (DST) is used
- This algorithm views bridges as nodes in
a graph and imposes a tree on the graph
(a tree is a graph that does not contain
cycles)
Loop is Removed
A
A
B
B
E
Loop Exists
E
Find routing from Eà F
EàBàAàCàF
EàAàCàF (lower hop count)
Another Example of Spanning Tree
Creating Bridge Table
Catalyst C
Blocked
Spanning Tree Protocol Variations
o 
Different variations of STP have been standardized
n 
n 
o 
IEEE standard 802.1Q provides a way to run STP on a set of
logically independent networks (VLAN)
n 
o 
o 
IEEE created a standard named 802.1D (in 1990)
the standard was updated in 1998
that share a physical medium without any confusion or interference
Cisco created a proprietary version of STP, Per-VLAN
Spanning Tree (PVST) for use on a VLAN switch
IEEE standard 802.1W introduced the Rapid STP (RSTP) has
been incorporated in 801.1d-2004 (in 1998), and now
replaces STP, some versions are
n 
n 
Multiple Instance STP (MISTP)
Multiple STP (MSTP)
VLAN Switches
Two LANs = Two hubs!
o 
Thus far, k LANs requires k hubs
n 
n 
o 
One solution is to establish Virtual Local Area Network (VLAN)
n 
o 
This can cause low efficiency; no load balancing
What if users are dispersed
Use VLAN switches; Make is flexible to add/remove users to each LAN
The concept of VLAN is straightforward:
n 
n 
n 
Allow a manager to configure a single switch to emulate multiple
independent switches
A manager can specify a set of ports on the switch and designates them to be
on virtual LAN 1, then designates another set of ports to be on virtual LAN 2,
and so on
When a computer on virtual LAN 2 broadcasts a packet only those computers
on the same virtual LAN receive a copy (i.e., once configured, a VLAN
switch makes it appear that there are multiple switches)
VLAN Switches
o 
o 
o 
o 
All Blues Computers
are in one LAN
Dividing computers into separate broadcast
domains is important
In each case, it may be important to guarantee
that a set of computers can communicate
without others receiving the packets and
without receiving packets from outsiders
In such cases, packets from each domain (LAN segment) has a separate color
Frame Colors can be determined different ways
n 
n 
n 
Each port of the Bridge (switch) is dedicated to a particular color à all domains
must be connected to the same VLAN
Each MAC address is recognized as a particular color à all domains must be
within the same organization
Each IP address is recognized as a particular color à the switch must act as level
3 device; mixing level 2 and 3 operations!
VLAN Example
Using Switches or Bridge Devices
A Practical VLAN Setup
VLAN
802.1.QFrame
VLAN
o 
o 
o 
o 
Max Total Bytes = 1518
CRC
Recomputed!
In 1988 802.1Q was established
Traditional Frame
n 
VLAN compliant Ethernet à changing the
frame max. length of Ethernet from 1518 to
46-1500 Data
1522 bytes
n 
Adding VLAN tag to each frame
PRI bits are used for QoS and supporting realtime applications
VLAN Frame
CFI is Canonical Format Indicator (Corporate Ego
Indicator!) – nothing to do with Ethernet
46-1500 Data
In case 802.1Q compliant switches are connected
to the traditional,
n 
802.1Q encapsulation inserts a 4-byte tag
field into the original Ethernet frame between
the source address and type/length fields and
Bits
re-computes the frame check sequence (FCS)
USR PRI 3
on the modified frame.
n 
The added 4 bytes are removed when the
frame is sent to a non-802.1Q node à next
slide
We can always add the VLAN field! (but not supported by current 802.3 devices!
http://routemyworld.com/2009/04/22/bcmsn-vlan-configuration-and-implementation/
SNAP: Sub-Network Attachment Point
The IEEE 802.1Q Standard
802.1Q bridges Plug & Pay!
Sources
o 
o 
o 
o 
o 
o 
Tomasi Text Book
Wireless lan
http://www.cisco.com/warp/public/cc/pd/witc/ao1200ap/prodlit/wswpf_wp.htm
LAN Design : http://module42k5.tripod.com/toyota.htm
Read about VLAN http://www.3com.com/other/pdfs/solutions/en_US/20037401.pdf
Tanenbaum Web resources
http://authors.phptr.com/tanenbaumcn4/webResources/
coverPageWebResources.html#VLAN
Network Efficiency
n 
http://www.erg.abdn.ac.uk/users/gorry/course/lan-pages/enet-calc.html
Projects
o 
Creating VLAN using Linux
n 
n 
http://www.candelatech.com/~greear/vlan.html
http://vimeo.com/6828914
Extra
Interconnection Between Devices
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

advertisement

Languages