Distributed Quality-Lifetime Maximization in Wireless Video Sensor Networks

Owing to the availability of low-cost and low-power CMOS cameras, wireless video sensor networks (WVSN) has recently become a reality. However video encoding is still a costly process for energy and capacity constrained sensor nodes and its optimal joint control with the communication protocols has a direct impact on the network lifetime. In this paper we propose a distributed quality-lifetime control algorithm where quality is simply measured by the visual signal quality. In order to formulate the quality-lifetime problem, we consider the power-rate-distortion (P-R-D) model of the video encoder together with the rate control, medium access and routing functions of the underlying communication protocol and formulate the problem as a generalized network utility maximization (GNUM) problem. Then we construct a distributed solution based on duality and proximal point methods with necessary convergence analysis. Simulation results support that optimal quality-lifetime control is possible through the proposed distributed algorithm, where the desired point of operation is simply adjusted by the sink via a configuration parameter.