On The Downlink Transmission Techniques Of Multi-user MISO/MIMO Wireless Communication Systems

Multi-user Multiple Input Single Output (MISO)/Multiple Input Multiple Output (MIMO) technology is a promising key technique for future wireless communication systems due to its potentials to achieve spatial multiplexing of multiple data streams and users at the same time and frequency channel as well as rich spatial diversity gain and multi-user diversity gain. Based on the existing worldwide innovative researches, this dissertation further investigates the downlink transmission techniques of multi-user MISO/MIMO wireless communication systems. Key contents of this dissertation include sub-optimal Dirty Paper Coding (DPC) scheme, linear transmission schemes based on Sum Mean Square Error (SMSE) minimization, robust linear transmission schemes and beamforming schemes in the limited-feedback multi-user MISO/MIMO downlink.In order to avoid the relatively high complexity introduced by calculating the optimal transmit processing matrices in the optimal DPC scheme, a sub-optimal Block Zero-Forcing DPC (Block ZF-DPC), associated with two multi-user scheduling algorithms, are proposed for the multi-user MIMO downlink. The Block ZF-DPC adopts sub-optimal transmit processing matrices calculated by simple matrix null space operations. The Block ZF-DPC scheme can asymptotically achieve the sum rate of the optimal DPC scheme if the number of users equals to the upper bound on the active users. Meanwhile, two sub-optimal multi-user scheduling algorithms which have linear complexity with respect to the number of users are designed for the Block ZF-DPC. Both sub-optimal scheduling algorithms can approach the sum rate of the optimal scheduling algorithm and can achieve the optimal spatial multiplexing gain as well as the optimal multi-user diversity gainAssuming that the perfect Channel State Information (CSI) is available at Base Station (BS), this dissertation investigates in detail the linear transmission schemes in the multi-user MISO/MIMO downlink and a novel linear transmission scheme based on SMSE minimization is proposed for the multi-user MIMO downlink. This novel scheme incorporates the total transmit power constraint into the objective function and solves the SMSE minimization problem by alternative optimizing the transmit/receive processing matrices in the downlink. Simulation results demonstrate that the SMSE minimization algorithm of this scheme is globally optimal and this novel scheme has better Symbol Error Rate(SER) performance,less complexity and faster convergence speed than the existing scheme.With the objective to mitigate the performance degradation induced by channel estimation error in the multi-user MIMO downlink, a robust linear transmission scheme based on conditional SMSE minimization is proposed. The distribution of the real channel conditional on imperfect channel estimates is firstly derived, then the transmit/receive processing matrices are optimized by minimizing the conditional SMSE under individual power constraints. Compared to the non-robust transmission scheme, this scheme exhibits obvious robustness to the channel estimation error, excellent convergence property and low extra computation load.The existing Zero Forcing Beamforming (ZFBF) scheme in the limited-feedback multi-user MISO downlink has a drawback that the total feedback amount increases linearly with the number of users. Based on the existing ZFBF scheme, a ZFBF scheme with feedback threshold is proposed. By introducing an appropriate feedback condition and a feedback threshold at each terminal, this scheme constrains the average total feedback amount by a relatively small and fixed upper bound. Based on the derived upper bound on the feedback probability and feedback threshold design criterion, the formula of the feedback threshold and an approximate method to compute the overflowing probability are presented and the asymptotical performance of this scheme is theoretically analyzed. Simulation results validate the threshold design and theoretical analysis and also demonstrate that the proposed ZFBF scheme with feedback threshold can approach the performance of the existing ZFBF scheme without feedback threshold with reduced total feedback amount.In the limited-feedback multi-user MIMO-Orthogonal Frequency Division Multiple Access (OFDMA) downlink, the existing Random Beamforming (RBF) scheme also has the drawback that the total feedback amount increases linearly with the number of users. Based on the existing RBF scheme, a RBF scheme with feedback threshold is proposed. This scheme introduces an appropriate feedback condition and a feedback threshold at each subcarrier so that the growing speed of the total average feedback amount with the number of users is slower than the natural logarithm function. According to the derived upper bound on the feedback probability and feedback threshold design criterion, the equation of the feedback threshold is derived. Meanwhile, the asymptotical performance of this scheme is theoretically analyzed and an upper bound on the overflowing bound is presented. Simulation results validate the threshold design and theoretical analysis and also demonstrate that the proposed scheme can asymptotically achieve the spectrum efficiency of the existing RBF scheme without feedback threshold and the optimal multi-user diversity gain as the number of users approaches infinity.