Single carrier frequency domain equalization and energy efficiency optimization for MIMO cognitive radio.

This dissertation studies two separate topics in wireless communication systems. One topic focuses on the Single Carrier Frequency Domain Equalization (SC-FDE), which is a promising technique to mitigate the multipath effect in the broadband wireless communication. Another topic targets on the energy efficiency optimization in a multiple input multiple output (MIMO) cognitive radio network. For SC-FDE, the conventional linear receivers suffer from the noise amplification in deep fading channel. To overcome this, a fractional spaced frequency domain (FSFD) receiver based on frequency domain oversampling (FDO) is proposed for SC-FDE to improve the performance of the linear receiver under deep fading channels. By properly designing the guard interval, a larger sized Discrete Fourier Transform (DFT) is equipped to oversample the received signal in frequency domain. Thus, the effect of frequency-selective fading can still be eliminated by a one-tap frequency domain filter. Two types of FSFD receivers are proposed based on the least square (LS) and minimum mean square error (MMSE) criterion. Both the semi-analytical analysis and simulation results are given to evaluate the performance of the proposed receivers. Another challenge in SC-FDE is the in-phase/quadrature phase (IQ) imbalance caused by unideal radio frequency (RF) front-end at the transmitter or the receiver. Most existing works in single carrier transmission employ linear compensation methods, such as LS and MMSE, to combat the interference caused by IQ imbalance. Actually, for single carrier transmissions, it is possible for the receivers to adopt advanced nonlinear compensation methods to improve the system performance under IQ imbalance. For such purpose, an iterative decision feedback receiver is proposed to compensate the IQ imbalance caused by unideal RF front-end in SC-FDE. Numerical results show that the proposed iterative IQ imbalance compensation can significantly improve the performance of SC-FDE system under IQ imbalance compared with the conventional linear method. For the energy efficiency optimization in the MIMO cognitive radio network, multiple secondary users (SUs) coexisting with a primary user (PU) adjust their antenna radiation patterns and power allocations to achieve energy-efficient transmission. The optimization problems are formulated to maximize the energy efficiency of a cognitive radio network in both distributed and centralized point of views. Also, constraints on the transmission power and the interference to PU are introduced to protect the PU’s transmission. In order to solve the non-convex optimization problems, convex relaxations are used to transform them into equivalent problems with better tractability. Then three optimization algorithms are proposed to find the energy-efficient transmission strategies. Simulation results show that the proposed energy-efficiency optimization algorithms outperform the existing algorithms