Widespread use of computer networks and the use of varied technolgy for the interconnection of computers has made congestion a significant problem. In this report, we summarize our research on congestion avoidance. We compare the concept of congestion avoidance with that of congestion control. Briefly, congestion control is a recovery mechanism. While congestion avoidance is a prevention mechanism. A congestion control scheme helps the network to recover from the congestion state while a congestion avoidance scheme allows a network to opperate in the region of low delay and high throughtput with minimal queuing, thereby, preventing it from entering the congestioed state in which packets are lost due to buffer shortage.

A number of possible alternatives for congestion avoidance were identified. From these alternatives we selected one called the binary feedback scheme in which the network uses a single bit in the network layer header to feedback the congestion information to its users, which then increase or decrease thier load to make optimal use of the resources. The concept of global optimality in a distribute system is defined in terms of efficiency and fariness such that they can be independently quantified and apply to thy number of resources and users.

The proposed scheme has been simulated and shown to be globally efficient, fair, responsive, convergent, robust, ditrisbuted, and configuration-independent.

Complete Report in Adobe Acrobat

A version of this report was published later in SIGCOMM and was selected as the best paper for publication in ACM transactions.

It was then republished in 25th anniversary issue of Computer Communications. and is a winner of SIGCOMM Test of Time Paper Award