Maximum Euclidean distance network coded modulation for asymmetric decode-and-forward two-way relaying

Network coding (NC) compresses two traffic flows with the aid of low-complexity algebraic operations, hence holds the potential of significantly improving both the efficiency of wireless two-way relaying, where each receiver is collocated with a transmitter and hence has prior knowledge of the message intended for the distant receiver. In this contribution, network coded modulation (NCM) is proposed for jointly performing NC and modulation. As in classic coded modulation, the Euclidean distance between the symbols is maximised, hence the symbol error probability is minimised. Specifically, the authors first propose set-partitioning-based NCM as an universal concept which can be combined with arbitrary constellations. Then the authors conceive practical phase-shift keying/quadrature amplitude modulation (PSK/QAM) NCM schemes, referred to as network coded PSK/QAM, based on modulo addition of the normalised phase/amplitude. To achieve a spatial diversity gain at a low complexity, a NC oriented maximum ratio combining scheme is proposed for combining the network coded signal and the original signal of the source. An adaptive NCM is also proposed to maximise the throughput while guaranteeing a target bit error probability (BEP). Both theoretical performance analysis and simulations demonstrate that the proposed NCM can achieve at least 3 dB signal-to-noise ratio gain and two times diversity gain