Binary Source Estimation Using a Two-Tiered Wireless Sensor Network

摘要

Binary source estimation using indirect observation by a cluster of wireless sensors in a two-tiered architecture is considered. The high correlation among sensors' observation is exploited in order to implement an efficient distributed Joint Source Channel Coding (DJSCC) scheme based on Parallel Concatenated Convolutional Codes (PCCC). A novel approach employing two sink nodes per cluster providing a Distributed Space-Time Block Coding (DSTBC) is proposed to get diversity gain over fading channels without need to set up an extra link between two sinks. The proposed iterative decoding algorithm is easily scalable to a large number of sensors and outperforms independent RSC decoders followed by a Maximum Likelihood (ML) detector.