Optimum power and rate allocation in video sensor networks

In a sensor network, each sensor has a limited energy supply. Therefore, it is critical to minimize the power consumed by each sensor to maximize its lifetime. Video sensor networks differ from conventional sensor networks in the fact that video compression at the sensor node consumes a significant amount of power comparable to that used for communication. This poses new challenges, and renders power efficient algorithms developed for sensor networks not suitable for video sensor networks. In this paper, we develop an algorithm for the minimization of the total consumed power by jointly optimizing the encoding power, the transmission power, and the source rate at each sensor node. Furthermore, MAC layer resource allocation is incorporated into the minimization problem. Video signal distortion due to compression, and packet losses in the wireless channel are studied as well. An efficient solution algorithm is developed using Lagrange duality to solve the minimization problem. Through numerical results, the trade-off between allocating power to the encoding process or the transmission process is characterized. Moreover, the interaction between the nodes competing for channel resources, and how this affects their power consumption and distortion levels is studied.