Binomial congestion control algorithms

Abstract

This paper introduces and analyzes a class of nonlinear congestion con- trol algorithms called binomial algorithms, motivated in part by the needs of streaming audio and video applications for which a drastic reduction in transmission rate upon each congestion indication (or loss) is problematic. Binomial algorithms generalize TCP-style additive-increase by increasing inversely proportional to a power of the current window (for TCP, ) ; they generalize TCP-style multiplicative-decrease by decreasing propor- tional to a power of the current window (for TCP, ). We show that there are an infinite number of deployable TCP-compatible binomial algo- rithms, those which satisfy , and that all binomial algorithms converge to fairness under a synchronized-feedback assumption provided . Our simulation results show that binomial algorithms interact well with TCP across a RED gateway. We focus on two particular algorithms, IIAD ( ) and SQRT ( ), showing that they are well-suited to applications that do not react well to large TCP-style window reductions. with the same con- stant of proportionality as for a TCP connection with the same packet size and round-trip time. In this paper, we present and evaluate a new class of nonlin- ear congestion control algorithms for Internet transport proto- cols and applications. Our work is motivated by two important goals. First, we seek to develop and analyze a family of algo- rithms for applications such as Internet audio and video that do not react well to drastic rate reductions because of the degrada- tions in user-perceived quality that result. Second, we seek to achieve a deeper understanding of TCP-compatible congestion control by generalizing the familiar class of linear control algo- rithms (of which AIMD is one example), and understanding how a TCP-compatible algorithm competes with TCP for bottleneck resources. While previous work on equation-based congestion control has shown how adjusting the transmission rate as a func- tion of the loss-rate enables interesting congestion control for streaming applications (12), our work opens up the possibility of using increase/decrease rules without tracking loss-rates. An AIMD control algorithm may be expressed as: I: