作者: Javad Haghighat , Hamid Behroozi , David V. Plant
DOI: 10.1109/VETECS.2009.5073550
关键词:
摘要: We consider a specific type of data gathering sensor networks that can be modeled by binary chief executive officer problem. apply turbo codes to encode sensors observa- tions and transmit them fusion center over independent symmetric channels. It is shown in the literature exploit correlation between observations design soft-input soft-output (SISO) global decoder. Then iterates extrinsic information decoder SISO applied error correcting code jointly estimate source. Since we codes, joint decoding problem generalized exchanging three modules. In this paper, first sum- product algorithm derive rules update for Then, extended- serial best known structure decoders consisting concatenated compare bit rate achieved extended-serial with one separate strategy, where separately decodes each sensor's observation then decides based on majority outputs. Our simulations show performs significantly better than decoding.