Wagner-like decoding for noncoherent PPM based ultra-low-power communications

Noncoherent pulse-position modulation (PPM) with simple channel codes has the potential to realize ultra-low power (ULP) wireless design. In this paper, we develop a Wagnerlike decoding rule for single-parity-check and high-rate Reed-Solomon (RS) coded PPM schemes by simply `flipping' the most unreliable received PPM symbol(s) to obtain a good balance between performance and coding complexity. The proposed algorithm can be considered as a list decoding algorithm that first generates a candidate codeword list based on the algebraic structure of the code before applying soft decisions to decode. This approach can result in more power-efficient realizations of the studied schemes. It is shown that our decoding approach can achieve near maximum likelihood decoding performance based on the trellis, while having a significantly lower decoding complexity. In addition, by exploiting the inherent advantage of PPM transmission, it is possible to reduce the candidate list to further simplify the decoding for RS-coded PPM without losing coding gain. This makes the proposed scheme more attractive for ULP communications.