T1: A Survival Guide

Chapter 10. Frame Relay

The great artist is the simplifier.

—Henri Frederique Amiel

Development of digital-transmission facilities in the 1960s and 1970s made low error rates a notable attribute of the telecommunications landscape in the 1980s and 1990s. Early packet-switching standards, most notably X.25, compensated for high error rates with extensive buffering and retransmission. Not all protocols that were run on X.25 links had reliability features, so providing reliability in the access protocol was a good thing. As the world moved toward protocols designed with the OSI model in mind, however, reliability and retransmission moved to higher layers of the stack. This resulted in duplicated effort and needless delays, especially on relatively clean digital links.

Frame relay was developed in response to the changing environment and, in many circles, was widely viewed as a direct replacement for X.25 networks. Frame relay specifies only the link layer interconnection of networks. Reliability is minimally addressed through the use of a frame check sequence that allows a frame relay network to discard corrupted frames; sequencing, data-representation errors, and flow control are delegated to higher-level protocols. (For the purposes of this book, most of the “higher-level” flow control related functions are delegated to TCP.)

To view frame relay strictly as a link layer takes too narrow a view; taken as a whole, frame relay provides the standards for building a packet-switched ...