1. Chapter 11. Frame Relay Background Frame Relay Protocol Architecture
Frame Relay Call Control
User Data Transfer
Network Function
Congestion Control

2. Background Key features of X.25
not only determines the user-network interface but also influences the internal design of the network
Call control packets are carried on the same channel and the same virtual circuit as data packets  inband signaling
Multiplexing of virtual circuits takes place at layer 3
Both layer 2 and layer 3 include flow control and error control mechanisms

3. Background (cont) Goal of frame relay
eliminate much of the overhead that X.25 imposes on end user systems and on the packet-switching network
Difference between frame relay and X.25
Call control signaling is carried on a separate logical connection from user data
Multiplexing and switching of logical connections take place at layer 2
No hop-by-hop flow control and error control
End-to-end flow/error control are the responsibility of a higher layer

4. Packet switching v.s. Frame Relay14 12
X.25 5
Packet-Switching
(X.25) 3 6
X.25 4
X.25 13 11 16 1 2 15 9 8 7 10
Source
Destination
Frame Relay FR 3 2 6 FR 7 FR 1 8 5 4
Source
Destination

5. Frame Relay Protocol Architecture

6. User Plane Transfer of information of btw end users
Uses core function of LAPF
Frame delimiting, alignment
Frame multiplexing/demultiplexing
Inspection of the frame to ensure that
it consists of an integral number of octets
it is neither too long nor too short
Detection of transmission error
Congestion control function

7. Frame Relay v.s. X.25

8. Frame Relay v.s. X.25 (cont)

9. Frame Relay Call Control
Terminal equipment
Network equipment
Exchange termination
Frame Handler

10. Frame Relay Call Control (cont)

11. Frame Relay Connection
Support multiple connections over a single link
Data link connection
DLCI: data link connection identifier
DLCI = 0
dedicated to call control
4 message types: SETUP, CONNECT, RELEASE, RELEASE COMPLETE

12. Control Signaling (e.g.)

13. Frame Format (LAPF-core)

14. Frame Format (cont) Only one frame type
used for carrying user data
no control frames
Not possible to use inband signaling
since a logical connection can only carry user data
Not possible to perform flow/error control
since there are no sequence number
DLCI
serves the same function as the virtual circuit number in X.25

15. Network Function Routing based on DLCI Multiplex multiple
logical connections

16. Congestion Control

17. Congestion Control (cont)

18. Congestion Control (cont)
Objectives (ITU-I Recommendation I.370)
Minimize frame discard
Maintain an agreed quality of service
Minimize the possibility that one end user can monopolize network resources at the expense of other end users
Be simple to implement, and place little overhead
Create minimal additional network traffic
Distribute network resources fairly among end users
Limit spread of congestion
Minimize the variance in quality of service delivered to individual frame relay connections during congestion

19. Congestion Control (cont)
Implicit (delay, discard)
Policing
Backpressure
S/W
S/W
S/W
S/W
Source
Destination
Choke packet
Explicit (binary, credit, rate)

20. Congestion Control (cont)
Backpressure
On the basis of links or logical connections
X.25-based packet-switching networks
Neither FR nor ATM has any capability for restricting flow on a hop-by-hop basis
Choke packet
A control packet generated at a congested node and transmitted back to a source node to restrict traffic flow
E.g. ICMP Source Quench packet
The source quench message can be used to a router or host that must discard IP datagrams because of a full buffer

21. Congestion Control (cont)
Implicit Congestion Signaling
By transmission delay and discarded packets
Effective technique in connectionless or datagrams configurations, such as IP-based internet
TCP: detect increased delay and segment loss
Explicit Congestion Signaling
The network alerts end systems and end systems take steps to reduce the offered load
Backward and Forward directions
Binary, Credit based, and Rate based approaches

22. Congestion Control (cont)
Limited tools available to the frame relay
Joint responsibility of the network and the end users
The network is in the best position to monitor the degree of congestion
The end users are in the best position to control congestion by limiting the flow of traffic
When congestion becomes severe enough
The network is forced to discard frames
In a way that is fair to all users

23. FR Congestion Control Discard strategy Congestion avoidance
Requires some explicit signaling mechanisms
Congestion recovery
Dropped frames are reported by some higher layer software
Implicit signaling mechanism

24. FR Congestion Control (cont)

25. Traffic Rate management
Reason
The simplest way to cope with congestion for FR is to discard frames arbitrarily
Since there is no reward for restraint, the best strategy is to transmit frames as rapidly as possible → exacerbates the congestion problem
Introduce the concept of a committed information rate (CIR)

26. Traffic Rate management (cont)
Committed information rate (CIR)
Committed Burst Size (Bc)
Max. amount of data that the network agrees to transfer over a measurement interval T
Excess Burst Size (Be)
The maximum amount of data in excess of Bc that the network will attempt to transfer over T
T = Bc / CIR or Bc = CIR * T
S CIRi, j <= Access Ratej for channel j i

27. Traffic Rate management (cont)

28. Traffic Rate management (cont)

29. Leaky Bucket Algorithm
pkt

30. Congestion Avoidance With explicit signaling
Two points of view about congestion
congestion always occurred slowly and almost always in the network egress nodes
congestion grew very quickly in the internal nodes and required quick, decisive action to prevent network congestion
Two approaches
Forward network congestion avoidance
Backward network congestion avoidance

31. Forward Explicit Congestion Notification
FECN
The notification indicates that this frame, on this logical connection, has encountered congested resources
Receipt of FECN signals
It requires the user to notify its peer user of this connection to restrict its flow of frames. It must be done at a higher layer, such as transport layer

32. Backward Explicit Congestion Notification
BECN
The notification indicates that the frames transmitted by the user (being notified) on this logical connection may encounter congested resources
Receipt of BECN signals
The user simply reduces the rate at which frames are transmitted until the signal ceases

33. restrict the flow of frames until the signal ceases
FECN v.s. BECN
Notify the peer user to
restrict the flow of frames
congested FR FR FR
BECN
BECN
FECN
FECN
User reduces the rate
until the signal ceases

34. Congestion Recovery With implicit signaling
Occurs when the network discards a frame, and this fact is detected by the end user at a higher, end-to-end layer
When this occurs, the end user software may deduce that congestion exists
Once congestion is detected
The protocol uses flow control to recover from the congestion