University of Bristol
Video Sensor Networks (VSNs) offer an alternative to several existing surveillance technologies. However, unlike in conventional sensor network, video processing and transmission requires large amount of resources both in signal processing, i.e., encoding, and transmission of the encoded data. For such networks, an optimal encoding power and rate allocation method based on a Power-Rate-Distortion (PRD) analysis has previously been proposed, where the power consumption of video encoding can be controlled by managing some encoding parameters. However, these parameters are currently obtained offline by examining the stored video, an approach which may not be suitable for surveillance applications.