Research On Synchronization And Omnidirectional Transmission In Massive MIMO Wireless Systems

With the booming development of Mobile Internet, more and more services will be provided to cus-tomers using mobile communication systems. Considering the severe shortage of wireless resources, the next generation mobile communication will no doubt pursue higher spectrum and power efficiency. Massive MIMO is a quite promising transmission technology with the spatial resource deeply exploited. In massive MIMO systems, directional beamforming is mainly adopted to transmit user-specific data. However, the signals transmitted in public channels should realize omnidirectional transmission in order to ensure reliable cell-wide coverage. This thesis focuses on the omnidirectional transmission technology for massive MIMO systems. Downlink synchronization and transmission scheme for public control signals are investigated.Firstly, omnidirectional precoding based transmission is carefully investigated, of which the precoding matrix design is critical. Combined with the channel characteristics of massive MIMO systems, three neces-sary conditions are deduced for the precoding matrix, in order to meet the requirements for the omnidirectional transmission, the sufficient utilization of power amplifiers and the ergodic mutual information maximization with the i.i.d. channel. Several examples of the precoding matrix are provided based on the Zadoff-Chu se-quence and the Golay sequence. By comparing cell-wide coverage, the ergodic mutual information and the peak-to-average power ratio corresponding to these designs, we pick out the designs which are more suitable for data transmission.Subsequently, the downlink synchronization for massive MIMO systems is investigated. As the key sig-nal for the cell search procedure, synchronization signal must be transmitted omnidirectionally. After giving a brief introduction of primary synchronization in LTE system, we point out the existing problems when classic transmission schemes are applied in massive MIMO systems. A new transmission schemes is proposed for massive MIMO systems, which uses Zadoff-Chu sequence as a transmitting weight vector. Furthermore, we point out that the transmitting signal can be generated with two synchronization sequences to acquire better cell-wide coverage performance. Based on maximum-likelihood criterion, the joint timing and frequency offset estimation algorithm is deduced. Considering the computational complexity and the restriction of ini-tial synchronization, we simplify it into M-part cross-correlation algorithm. Simulation results reveal that the synchronization signal generated by two synchronization sequences can provide better synchronization performance.Finally, we focus on the transmission of the public control signals such as broadcasting signals and control signals. With the omnidirectional precoding based transmission discussed above, we put forward an easily applied transmission scheme which can significantly reduce the downlink pilot overhead and provide transmission diversity. Taking inter-cell interference into consideration, we propose a receiving algorithm with inter-cell interference cancelled iteratively. The interference signals are detected, decoded and recon-structed by the receiver. Consequently the interference signals are well eliminated from the received signal. Simulation results show that the proposed scheme can not only guarantee the omnidirectional transmission but also improve the robustness of the transmission. Moreover, the inter-cell interference cancellation algorithm is quite efficient in suppressing strong inter-cell interference.