PSNR-FAIR STATISTICAL MULTIPLEXING WITH DISTRIBUTED CONTROL OF VIDEO SERVERS

This paper proposes a partly distributed statistical multiplexing (SM) system able to meet some video quality fairness target among programs without requiring exchange of information between servers. The transmission rate allocation among programs is performed by some Media Aware Network Element (MANE). The MANE feeds back to each video server the level of its associated buffer as well as a reference level. Each video server then adjusts the rate-distortion trade-off of the content it delivers so that its buffer in the MANE reaches the reference level. The transmission and encoding rate control problems are addressed with a control-theoretic approach. The system is described by a multi-input multi-output model and several Proportional Integral (PI) controllers are used to adjust the received video quality as well as the buffer levels. The study of the system equilibrium and stability provides guidelines for choosing the parameters of the PI controllers. Experimental results show that for video sources, quality fairness, and buffer level constraints are satisfied. Index Terms-- Video coding, Multiplexing, Distributed control, Proportional-Integral control I. INTRODUCTION Statistical Multiplexing (SM) [1] of compressed video streams consists in adapting jointly the rate and the quality of the multiplexed videos in order to use the available bandwidth in an optimal way. Usually, Variable Bit Rate (VBR) [2] video coders are used in order to perform this adaptation. When performing SM, users should receive decoded contents with (objective or subjective) quality of the same order of magnitude (quality fairness).