Title: Frame Relay
1Frame Relay ATM
2Virtual Circuit Background
- In the beginning there was X.25
- Then came Frame Relay and ATM
3Virtual Circuit Background X.25
- X.25
- 1970's technology
- Implemented at the network layer
- 64 kbps
- Implements extensive error correction and flow
control.
4Virtual Circuit Background Frame Relay
- Frame Relay
- Late 1980's 1990's technology
- Operates at rates ranging from 1.544 Mbps (DS-1
T1) to 44.376 Mbps (DS-3 T3). - Operates at the physical and network link layer
so it can interface with existing Network layer
protocols. - Error detection at the data link layer. No flow
control or error correction control (less
overhead) - Allows for digital voice communication (VOFR)
5Frame Relay
- ATT is the world's largest provider of Frame
Relay service.
6Frame Relay Virtual Circuits
- Each switch in a frame relay network has a
forwarding table. - Virtual circuit identifiers in frame relay are
dubbed as data link connection identifiers
(DLCI).
7Figure 8.13 Source-to-destination data transfer
in a virtual-circuit network
8Figure 18.1 Frame Relay network
9Figure 18.2 Frame Relay layers
10Figure 18.3 Frame Relay frame
11Frame Relay Frame
- The flag at the start of the frame has bits
- 01111110
- 16 bit Address information field
- Data field (up to 9000 bytes)
- FCS error detection using CRC.
12Address Field
- The address field contains sub-fields that allow
for congestion control (quality of service) - QoS allows a frame relay switch to drop some
packets if there is data congestion. - DE is the discard eligibility field. If this bit
is set, the switch can discard the frame during
congestion.
13ATM Asynchronous Transfer Mode
- The combination of ATM and SONET allows
high-speed interconnection of all the world's
networks. - ATM is a virtual switching technology
- SONET is a transmission technology
14ATM
- Designed in the 1980's
- Designed to take advantage of the bandwidth of
optical fiber transmission media. - Many of the protocol functions are implemented in
hardware (not software) to insure the best
possible performance. - 1.54Mbps to 155 Mbps user copper, or 622 Mbps
using optical fiber.
15Background Frame Size vs Performance
- Each layer of a protocol adds header information.
- The ratio of data to packet size diminishes as
more protocol layers are added. - Increasing the data payload leads to
inefficiencies in data transmission - Small packets are padded
- Large packets may need retransmission wasting
bandwidth - Large packets delay other packets
16Frame Size vs Multiplexing
- A low priority large frame can delay higher
priority small frames if the large frame arrives
before the small frames.
17Figure 18.6 Multiplexing using different frame
sizes
18Solution Cell Networking
- The single uniform size cell makes transmission
of data more predictable. - No frame size issues exist when multiplexing same
size cells. - Small cells arriving at rapid speed can appear to
be nearly continuous. - Voice video
19Figure 18.7 Multiplexing using cells
20ATM Multiplexing
- ATM uses asynchronous time division multiplexing.
21Figure 18.8 ATM multiplexing
22ATM Multiplexing
- A time slot in an asynchronous ATM multiplexed
link will be empty only if there is no data from
any of the input links.
23ATM Architecture
- Cell switched architecture.
- UNI user to network interfaces
- NNI network to network interfaces
24Figure 18.9 Architecture of an ATM network
25Figure 8.13 Source-to-destination data transfer
in a virtual-circuit network
26ATM Cell Format
- All cells are 53 bytes
- 5 byte header
- 48 byte data payload
27Figure 18.14 An ATM cell
28Figure 18.11 Example of VPs and VCs
29VCI, VPI
- The switches connected to user interfaces and
network interfaces rely on both the virtual
circuit identifier and virtual path identifier - The NNI relies on the VPI for forwarding.
- The UNI relies on both the VCI VPI
30ATM Cell Header Format
- The 5 byte header contains VPI and VCI
information. - Both the VPI and VCI are used to forward
information to its destination.
31Figure 18.13 Virtual connection identifiers in
UNIs and NNIs
32Figure 18.12 Connection identifiers
33ATM Virtual Switching
- The routing tables have six fields
- Input
- Interface number
- VPI
- VCI
- Output
- Interface number
- VPI
- VCI
34Figure 18.15 Routing with a switch
35ATM Layers
- ATM is implemented in the physical and data link
layers. - Any physical layer carrier can carry ATM cells
(wired, optical, wireless).
36Figure 18.17 ATM layers in endpoint devices and
switches
37The ATM Layer
- The ATM layer provides
- Routing
- Switching
- Multiplexing
- Flow control (quality of service)
- Error detection and correction
38ATM's AAL Application Adaptation Layer
- There are four variations of AAL to deal with
different data types. - AAL-n is designed to accept
- AAL1 handles data streams of audio (voice) and
video by allocating the streams into cells. - AAL-5 handles other data communications like
Internet traffic.
39Commercial ATM
- ATM is used in DSL and B-ISDN services (Digital
Subscriber Lines and Broadband Integrated
Services Digital Network) - ATM works with any transport medium and ATM
streams are multiplexed into SONET streams.