Reliable signal transmission in wireless sensor networks with zero bandwidth expansion

This paper considers wireless sensor networks with limited bandwidth, where sensors must transmit sensing results that are distorted with noise uncertainty, to the fusion center reliably and efficiently. A class of rate-1 linear complex-field coding (CFC) transmission strategy is developed, to effectively protect the noisy observation of the signals without any bandwidth expansion. Pairwise error probability (PEP) is analyzed, design criterion for optimal complex-field coding strategies is derived, and optimal codes are examined. In addition to maximum likelihood (ML) decoding, a low-complexity decoding scheme termed partially nulling and canceling (PNC) is also proposed. The PNC decoder, representing a midway tradeoff between sphere decoding and (linear) minimum mean square error (MMSE) detection, strikes a good balance between complexity and performance. Extensive simulations show that the proposed CFC is able to reap considerable gains without any bandwith epansion!