Distributed Vector Ouantization Over Sensor Network

Vector quantization is an algorithm to encoding the original data to reduce the bits needed for communication and storage saving while maintaining the necessary fidelity of the input data. LBG algorithm and SOM algorithm are two popular vector quantization algorithms. Signal processing over distributed network has received a lot of attention in recent years. Gathering data to a central processing node is usually infeasible for sensor network due to limited communication resource and power. In distributed algorithm, each sensor node deals with its own input data and transmits the local results to its neighbors. Each node fuses the information from neighbors and its own to get the final results. The local results instead of the original data are transmitted so that the communication complexity is reduced. In this paper, we consider the distributed vector quantization problem and develop distributed LBG algorithm, distributed SOM algorithm and distributed vector quantization based on Kullback-Leibler divergence, based on traditional LBG algorithm, traditional SOM algorithm and centralized vector quantization algorithm based on Kullback-Leibler divergence. Numerical simulation shows that our distributed algorithms can reduce the communication complexity in distributed vector quantization problem.