Design and implementation of partially-coherent architectures for QAM communications.

摘要

The fast-growing need for high data rate transmission in wireless environment has sparked interest in the high level modulation schemes, such as M-ary Quadrature Amplitude Modulation (M-QAM), which can offer a trade-off between performance (or quality of service (QoS)) and data rate. It is also well known that coherent detection mechanism is used widely for QAM in the emerging wireless standards. However, the performance of coherent detection is highly sensitive to the accuracy of the phase estimation and tracking process.;In this dissertation, we consider a coherent environment, where phase synchronization is established via a phase-locked loop (PLL) mechanism. It is assumed that the resulting phase estimation error obeys Tikhonov distribution. Provided such a residual phase error, we extend the result in the literature to the case of QAM modulation and obtain a partially-coherent receiver for the detection of QAM signals using maximum a posteriority (MAP) rule. Provided the complexity of the resulting optimum receiver, we propose a computationally-efficient architecture.;The performances of the proposed architectures are evaluated in terms of bit error rate (BER) using simulation as well as approximate analytical methods. Two different QAM modulation schemes, i.e., modulations with 8- and 16-ary constellations, are studied. In both scenarios, the performances of the proposed computationally-efficient receiver are shown to follow those of the optimum receiver over a wide range of signal-to-noise ratios (SNR), while outperforming a standard coherent receiver operating in the presence of residual phase error by as much as 2 dB in AWGN channel. The complexity of the computationally-efficient receiver architecture is evaluated via implementing the proposed architectures on an FPGA platform.