Low-cost implementation of M-QAM signal constellations

Abstract

This paper aims to discuss ways of reducing hardware resource usage required for implementing vectorial coding and detection of a variable-order M-ary quadrature amplitude modulation (M-QAM) constellations in digital communication systems. In this regard, we introduce two techniques that are capable of mapping both square and cross M-QAM constellation symbols to bits and vice-versa, with distinct levels of hardware resource utilizations when M increases. These techniques are named Generic M-QAM and Extended Heuristic Decision Region (HDR). The latter technique modulates, demodulates, and detects M-QAM symbols with look-up tables and state machines, while the former technique performs the same tasks by using few logical elements and eliminating memory usage. Based on additive white Gaussian noise model assumption and hard decision rule, we show the bit error rate performances obtained with both techniques are equal to theoretical ones. Additionally, the attained results show the Generic M-QAM technique offers the lowest hardware resource utilization when M increases (M>256) in comparison with the Extended HDR one. Moreover, we point out that the Extended HDR technique has the flexibility to be easily implemented in any published standards, while the Generic M-QAM technique applies a Gray mapping function that might require adjustment to make it compatible to published standards.