A non-cooperative quality optimization game for scalable video delivery over MANETs

Abstract

Non-cooperative game theory has found many important applications in engineering. In networking research community, this theory has been adopted to represent the selfish behaviour of the network users. In video transmission over Mobile Ad hoc NETworks (MANETs), video sources try to optimize their perceived video quality unilaterally. On the other hand, network providers try to set some prices on the scarce network resources such that their associated income is maximized. In the current paper, a non-cooperative quality game framework has been introduced in which by exploiting a related quality metric, scalable video users try to optimize their perceived quality in a distributed manner. The existence of a Nash Equilibrium (NE) has been proved and the stability of the distributed resource allocation strategy has been investigated. In the proposed resource assignment game, the utility functions of all competing scalable video sources converge to some optimal value which is dictated by the resulting Nash equilibrium in the allocated rates. In this optimal point, no player (video source) can enhance its utility unilaterally without compromising other players' utility. The assigned rates in the game can be considered as rate-feedbacks for on-line rate adaptation of a moderate scalable video encoder such as H.264/MPEG4 AVC. Some numerical analysis have been presented to validate the theoretical results and to verify the claims.