Using receiver-based rate matching for congestion detection in rate-based protocols

Traditional congestion control schemes rely solely on packet losses as an indication of congestion. This paper motivates and describes the use of receiver-based rate estimation for the proactive detection of incipient congestion in rate-based protocols. By matching the actual and expected receive-rates, congestion can often be detected before packet losses occur. We present an algorithm to implement this concept in the TCP-friendly rate control (TFRC) scheme. We utilize the algorithm in a modification to TFRC for the optimization of the session start-up phase. Simulation results show that this modification can help to avoid packet losses and to stabilize the transmission rate quicker at session start-up. We point out that the same principle can be used to avoid packet losses in other situations as well, such as the adaptation to rapidly changing link characteristics in wireless systems.