SenseVid : A traffic trace based tool for QoE Video transmission assessment dedicated to Wireless Video Sensor Networks

Video applications are being a key component to enhance traditional wireless sensor networks (WSN) applications. As a result, sensor network researchers need adequate and easy to use tools to assess the performances of their proposals. The existing transmission evaluation tools either consider video sequences along with codecs that are unsuitable for WSNs or make use of low cost compression methods for still images without inter-frame coding required for efficient video transmission. In this paper, we present SenseVid, an open source video transmission and evaluation tool that considers WSN specific characteristics. Besides low energy intra-frame compression based on fast pruned discrete cosine transform (DCT), a low complexity inter-frame encoding is provided to allow efficient support of video flows. A configurable fine-grain energy model is provided where both video capture and encoding cost are accounted for on a per frame basis. Video traffic differentiation based on priority levels is also provided. SenseVid adopts the video traffic traces approach, allowing its use in any simulation or real testbed environment. Using SenseVid, the user is able to reconstruct the received video considering lost packets during its transmission as well as estimating the achieved quality of service (QoS) and quality of experience (QoE).