Research On Blind Source Separation In Wireless Sensor Networks Based On Channel Coding

Wireless Sensor Networks have been widely applied in the fields of engineering andtechnology, such as safeguarding of military and public security and modern medical care,etc. Much attention has been paid to the research of distributed estimation of signal vectorsin WSN ever since. Blind source separation as a fundamental problem in signal processinghas been studied under the context of sensor arrays traditionally. Extending the problemfrom sensor arrays to wireless sensor networks is really a new research subject. Theinfluence of path loss and channel distortion makes it necessary to take the research of errorcontrolling into the problem of blind source separation in WSN.Low Density Parity Check (LDPC) codes have received intensive research for theirShannon capacity-approaching performance and high speed parallelizable decoderarchitecture. Considering its good performance and relatively low complexity, we proposedto use LDPC codes for error control in WSN. Because it has a stronger antinoise ability thanother ones, it can markedly improve the coding gains which makes the WSN adapt to amuch worse channel environment and enhance the network service survivability.In this paper, we first introduce the problem of blind source separation in WSN andthen take the channel coding into account, the principles of coding and decoding algorithmof BCH codes and LDPC codes are also presented. Based on the research, a lot ofsimulations are made to find out that the design of parameters have great influence on theperformance of LDPC codes which guide us to research on the influence it will have on theproblem of blind source separation in WSN. The fact is presented that in the condition ofrequest of certain bit error rate, LDPC codes perform better energy efficiency than BCHcodes in WSN. We then analysis the differences between the performance of using channelcoding and without channel coding in the certain problem of blind source separation inWSN. The simulation results show that using channel coding can greatly improve theperformance of blind source separation.