An approximately multi-user OFDM transceiver was introduced to reduce the multi-access interference (MAI ) due to the carrier frequency offset (CFO) to a negligible amount via precoding by Tsai, Lin and Kuo. In this work, we investigate the performance of this precoded multi-user (PMU) OFDM system in a time-variant channel environment. We analyze and compare the MAI effect caused by time-variant channels in the PMU-OFDM and the OFDMA systems. Generally speaking, the MAI effect consists of two parts. The first part is due to the loss of orthogonality among subchannels for all users while the second part is due to the CFO effect caused by the Doppler shift. Simulation results show that, although OFDMA outperforms the PMU-OFDM transceiver in a fast time-variant environment without CFO, PMU-OFDM outperforms OFDMA in a slow time-variant channel via the use of M/2 symmetric or anti-symmetric codewords of M Hadamard-Walsh codes.
Proc. SPIE. 5440, Digital Wireless Communications VI
KEYWORDS: Signal to noise ratio, Antennas, Orthogonal frequency division multiplexing, Optimization (mathematics), Receivers, Monte Carlo methods, Systems modeling, Signal attenuation, Wireless communications, Matrices
A maximum likelihood estimator (MLE) that jointly estimates the carrier frequency offset (CFO) and the channel response of each user in uplink OFDMA systems is investigated in this research. The proposed MLE distinguishes itself from existing methods by its applicability to more flexible carrier assignment schemes. It achieves high computational efficiency by transforming a multidimensional optimization problem into a one-dimensional optimization problem. A suboptimal method is developed to further reduce the computational complexity. It is demonstrated by simulation results that the proposed MLE can provide accurate CFO and channel estimation in both SISO and SIMO environments.
A symbol-by-symbol maximum likelihood (ML) detection scheme for multicarrier (MC) systems is proposed in this work. When the number of subchannels is sufficiently large, the received symbols across all subchannels are approximately uncorrelated. Then, the proposed symbol-by-symbol ML detection, which is obtained by minimizing the symbol error probability, is nearly optimal. Furthermore, we will show how to reduce the complexity of the symbol-by-symbol ML detection when the constellation size is large. Simulation results show that the proposed symbol-by-symbol ML detection scheme outperforms the symbol-by-symbol minimum distance (MD) detection scheme by up to 2 dB in a noisy environment with crosstalk such as the DMT-ADSL system.