List Slab-Sphere Decoding: Efficient High-Performance Decoding for Asymmetric MIMO Antenna Systems

Abstract

While capacity-achieving forward error-correcting codes (e.g., space-time turbo code) are available, the practical challenge, nonetheless, is the tremendous complexity required by joint optimal maximum-likelihood (ML) decoding. For this reason, iterative soft decoding has long been studied to approach the optimal ML decoding performance at practical complexity. In multiple-input multiple-output (MIMO) channels, a reasonable decoding strategy consists of two stages: (1) estimate the soft bits (i.e., the log-likelihood ratio (LLR) of each transmitted bit] using the list version of sphere decoding or its variants, and (2) update the soft bits through iterative soft decoding. A good MIMO decoder is required to produce reliable soft-bit estimates at the first stage before iterative soft decoding is performed. In this paper, we focus on the asymmetric MIMO antenna systems where the number of receive antennas is less than the number of signals multiplexed in space (or fat MIMO channel). We aim to design a low-complexity MIMO soft decoder that can achieve promising performance. The main idea of the proposed method is based on the generalization of sphere decoding for asymmetric fat MIMO channels (referred to as slab-sphere decoding). By having the list version of slab-sphere decoding, we can produce good soft-bit estimates at low complexity, thereby making efficient high-performance decoding possible.