An Information-Diversification and Phase-Offset-Precoding Method for Asynchronous Multiuser OFDM Systems

This paper investigates multiuser OFDM systems in which multiple users (transmitters) transmit signal to a common receiver at the same time. We assume the users adopt non-orthogonal signalling (i.e., the signals of different users are carried on all subcarriers of OFDM). We consider a simple, non-iterative, detector that works well under near-power balanced condition - the conventional SIC (successive interference cancellation) only works well under the power-imbalanced situation. For our detector, however, symbol misalignment and carrier frequency offsets (CFO) among packets of different users may lead to degradation of decoding performance. To overcome such performance degradation, we put forth the idea of spreading bit information across multiple adjacent subcarriers to achieve diversification. To ensure good BER performance, we realize this idea using an analytical-based precoding scheme at the transmitters, consisting of two parts: 1) a convolutional information-diversifying precoder, followed by 2) a phase-offset precoder. At the receiver side, we use a non-iterative belief propagation (BP) algorithm to optimally decode information from multiple users. Simulation results show that for a two-user BPSK (QPSK) system, our precoding method brings about 1.5dB (1.5 dB) BER gain compared with the non-precoded scheme.