Generalized Nearest Neighbor Decoding

It is well known that for Gaussian channels, a nearest neighbor decoding rule, which seeks the minimum Euclidean distance between codeword and received channel output vector, maximum likelihood solution hence capacity-achieving. Nearest remains convenient yet mismatched general key message of this paper performance can be improved by generalizing its metric to incorporate state dependent processing scaling. Using generalized mutual information, lower bound capacity under independent identically distributed codebook ensemble, as measure, establishes optimal input. Several restricted forms rule are also derived compared with existing solutions. The results illustrated through several case studies fading channels imperfect receiver information quantization effects.