13.3.2 Frame Relay

Frame Relay emerged during the late 1980s as a simplified, high-speed alternative to X.25. Rather than assuming unreliable transmission links, Frame Relay was designed for modern digital circuits with very low error rates. This allowed much of the error-control processing performed by X.25 to be eliminated.

Like X.25, Frame Relay provides connection-oriented communication using virtual circuits identified by Data Link Connection Identifiers (DLCIs). Users connect to the network through a Frame Relay Access Device (FRAD), which performs the necessary framing and interface functions.

Unlike X.25, however, Frame Relay performs only minimal processing within the network. Frames are checked for transmission errors, but if an error is detected the frame is simply discarded rather than retransmitted. Error recovery, flow control, and retransmission are instead performed by higher-layer protocols operating in the communicating devices themselves. This significantly reduces processing delay within the network and allows much higher transmission speeds.

Most Frame Relay services employed Permanent Virtual Circuits (PVCs) that remained continuously available, although switched virtual circuits were also defined. Because the network no longer performed extensive error recovery, Frame Relay switches were considerably simpler than X.25 switches and could forward traffic much more rapidly.

Carrier services were offered at a wide range of access speeds, including 56 kbps, 128 kbps, 256 kbps, 384 kbps, T1 (1.544 Mbps), E1 (2.048 Mbps), T3 (45 Mbps), and E3 (34 Mbps). Frame Relay became one of the dominant wide-area networking technologies throughout the 1990s, particularly for connecting geographically distributed corporate networks.

Although Frame Relay represented a significant improvement over X.25, it too was eventually superseded by technologies such as MPLS and high-speed Ethernet, which provided greater scalability, flexibility, and support for modern IP networking.