1.
MULTI PROTOCOL LABEL SWITCHING (MPLS)

2.
MPLS Overview
• A forwarding scheme designed to speed up IP
packet forwarding (RFC 3031)
• Idea: use a fixed length label in the packet
header to decide packet forwarding
– Label carried in an MPLS header between the link
layer header and network layer header
• Support any network layer protocol and link
layer protocol

3.
MPLS Header Format
• Label: 20-bit label value
• Exp: experimental use
– Can indicate class of service
• S: bottom of stack indicator
– 1 for the bottom label, 0 otherwise
• TTL: time to live
20 3 1 8
Label Exp S TTL

4.
Why MPLS Protection?
• IP restoration is very slow
– OSPF, RIP, etc. require a redistribution of
updated link status information in response to a
failure.
– Routing table convergence time on the order of
seconds.
– Looping and packet loss can occur during
convergence
• MPLS enables fast failure restoration

5.
MPLS Protection Approaches
• End-to-End protection
– A backup LSP is set up in advance from the source
LSR to the destination LSR of the primary LSP.
• The backup LSP is link and node disjoint with the primary
LSP
• Need reserve resources for the backup LSP
– Source LSR responsible for restorationsender must
be notified of the failure

6.
MPLS Protection Approaches
• Local protection
– When establishing a primary LSP, a backup LSP
for each possible link or node failure is set up
• Resources reserved for each backup LSP
– Failure detecting LSR responsible for switching
traffic to the backup LSR
– Faster restoration than end-to-end protection

7.
Local Protection
• Problem: must create a separate set of
backup LSPs for every primary LSP
• Can a single LSP backup a set of primary
LSPs?
• Yes! Use MPLS label stacking.

8.
Label Stacking
• A packet may carry multiple labels, organized as a last-
in-first-out stack
• A label may be added to/removed from the stack at any
LSR
• Processing always done on the top label
• Allow the aggregation of LSPs into a single LSP for a
portion of the route, creating a tunnel
– At the beginning of the tunnel, the LSR assigns the same label
to packets from different LSPs by pushing the label onto each
packet’s stack
– At the end of the tunnel, the LSR pops the top label

9.
Local Protection Using Label
Stacking
• Bypass tunnel: a LSP used to protect a set
of LSPs passing over a common facility.
• Label stacking allows different primary
LSPs to use the same bypass tunnel for
failure protection.

10.
Local Protection Using Label
Stacking
When a failure occurs:
• LSR at the beginning of the tunnel will
– Switch packets received on the protected LSP x onto the bypass
tunnel
– Replace the old label with a new label that will be understood by the
last node in the bypass tunnel to indicate LSP x
– Push the bypass tunnel's label onto the label-stack of the redirected
packets.
• LSR at the end of the tunnel will
– Pop the bypass tunnel's label
– Examine the top label to determine the protected LSP that the
packet is to follow.

11.
Frame Relay
What is it??
• Frame Relay (FR) - public network WAN technology based on
packet switching
• FR standard defines an interface between an end user and a
public network. FR is a protocol of 2nd level of OSI model
end user
Frame Relay cloud
Frame Relay
interface
Internal Frame Relay protocol (between switching devices in the cloud) is not
standardized (probably it will be some day)

12.
Frame Relay
What is it about?
• Aim: transport user data between port A and B
FRAD - Frame Relay
VC - Virtual Circuit
Access Device PVC - Permanent VC
PVC
A B
FR switches
• Data is transmitted as variable length frames
Max. frame length is 4096 bytes (recommended length is 1600
bytes)
• From users point of view: ports A and B are connected with a
transparent logical link (virtual circuit - VC)

13.
Frame Relay
Standards
• Frame Relay “independent existence”:
– In 1990 “Group of Four” (DEC, Northern Telecom, Cisco, Stratacom)
presented FR as an independent standard
– Later this Frame Relay Forum was established: main standardization body
for FR
• Standards on which FR is based:
ANSI T1.602, ANSI T1.606 (Frame Relaying Bearer Service - Architectural
Framework and Service Description, 1990), ANSI T1.607-1990, ANSI
T1S1/91-659,ANSI T1.617, ANSI T1.618, CCITT I.122 (Framework for providing
Additional Packet Mode Bearer Services, 1988), CCITT Q.922, CCITT Q.933

14.
Frame Relay
Most important features
• Based on packet (frame) switching
• Frames of variable length (up to 4096 bytes, typically 1600 bytes)
• Connection oriented; only permanent connections - PVCs; switched VCs in
standard extensions
• High data rates at user-network interfaces (2Mbps, ultimately up to 45 Mbps)
• Bandwidth on demand
• No flow control mechanisms (nearly)
• No error control (but FCS) or retransmission mechanisms
• All protocol functions implemented at 2nd level (data link) of OSI model
No standards for physical interface: can be X.21, V.35, G.703, G.704

15.
Frame Relay
Why was it proposed?
• Efficiency: increased demand for high throughput networking
(X.25 too slow)
• “Bursty applications”: LAN connectivity, Internet, not only
terminal applications
• Fibre optic lines: low (very, very low) bit error rates
• New, smarter software: applications (or higher level protocols
like TCP) performing error control, retransmissions; reliable date
links delivered by higher levels of OSI model

16.
Frame Relay
Topologies
• star
• full mesh

17.
Frame Relay
FR versus leased line
Advantages:
• Decreases number of ports on user devices
– important for star topology
– vital for full mesh topologies ( N(N-1)/2 connections, N(N-1) ports)
• Backup lines become public operator responsibility and no
longer that of an end user; backup connections are switched
transparently to the user
• More bandwidth is available for traffic peaks; CIR can be more
expensive than similar leased line; CIR+EIR is much cheaper

18.
Frame Relay
FR versus leased lines
Advantages:
• Allows to build virtual LANs over whole countries
(because of mesh topology and ARPs); simplifies routing
• Allows to build private virtual corporate networks; they
can be separated from the world at the 2nd level of OSI
model - safety
• A private network can be connected to the Internet in
only one point: safety and economy

19.
Frame Relay
FR versus leased lines
Advantages:
• Simplicity of the configuration for the end user
equipment (not necessarily for the operator…)
• Example: IP over Frame Relay on Cisco IOS
interface serial 0
ip address 194.1.1.1 255.255.255.0
encapsulation frame-relay ietf
frame-relay lmi-type ansi

20.
Frame Relay
FR versus leased lines
Disadvantages:
• Not for delay sensitive applications like: voice, video (though the
former is sometimes transmitted over FR)
• No guarantee that frames are delivered to the end point; is CIR
really CIR?
Lots depend on the FR operator; especially overbooking - how
many times sum of all CIRs extends physical capacity of
operators connections

21.
Frame Relay
How do you really use it
• Rent ports at the operator’s switches (normally together with
local leased lines and modems); you have to select clock rates
• Ask for PVCs between ports you want; it can be your ports,
ports on publicly available devices, like border router
• Configure your FRADs - see Cisco example
Isn’t it simple??

22.
ATM
• ATM standard (defined by CCITT) is widely accepted by
common carriers as mode of operation for
communication – particularly BISDN.
• ATM is a form of cell switching using small fixed-sized
packets.
Basic ATM Cell Format
5 Bytes 48 Bytes
Header Payload

23.
Asynchronous Transfer Mode
(ATM)
Voice
Data
packets MUX
Wasted bandwidth
Images
TDM
4 3 2 1 4 3 2 1 4 3 2 1
ATM `
4 3 1 3 2 2 1