A low-complexity soft QAM de-mapper based on first-order linear approximation

Abstract

High-order QAM modulation schemes are recommended in various communication systems. However, soft demodulating a high-order QAM constellation is complicated when applying traditional methods. In this paper, a novel de-mapper is proposed based on the fact that the clipped LLR can also offer sufficient soft-information, meanwhile the theoretical LLR curves are approximately linear in this clipped range. Hence a first-order linear function is designed to approximate the theoretical value. Simulation results and complexity analysis demonstrate that the proposed de-mapper outperforms traditional approaches in terms of BER performance, and also keeps low complexity.