Research On Optimization Of Non-linear Precoding In MIMO System

MIMO utilizes multiple antennas to provide spatial diversity and spatial multiplexing so as to obtain higher channel capacity, transmission reliability and spectral efficiency compared with traditional SISO system. Because of MIMO using multiple antennas for data transmission, the original data can be restored through complex detection techniques due to the inevitable interference existing among multiple data streams or users. However, client devices tend to be restricted on the size, power and other conditions. At the same time, the location of the client devices tend to be distributed, so it is difficult to use joint detection algorithm for client devices. Different from the MIMO detection performed at the receiver, the transmitter can use ideal CSI to process the sending data so as to suppress the multiuser interference in MIMO system.We call this technique as precoding algorithm. By using the precoding algorithm, all the data steams on the antennas can be transmitted in parallel without interference and channel capacity can be greatly increased. Additionally, precoding algorithm can greatly simplify the receiver process and reduce the system complexity.Precoding algorithm can be divided into linear precoding algorithm and nonlinear precoding algorithm. Linear precoding algorithm is simple, but the performance is poor. The complexity of nonlinear precoding algorithm is higher, but the performance is better than that of linear precoding.This thesis mainly discusses the nonlinear precoding algorithm with ideal CSI in MIMO system. This paper first discussed the basic principle of several main nonlinear precoding algorithms, namely, TH precoding, VP precoding, VP precoding based on LRA. In these three precoding algorithms, performance of VP precoding which is similar with ZF algorithm in coding structure is the best, but its complexity is high when determining the perturbation vector. Therefore, VP algorithm is difficult to apply. Later in this paper, In order to optimize the performance of VP precoding, ZF and VP jointed precoding algorithm is intensive studied under the condition of perfect CSI in multi-user MIMO system. Based on this algorithm, THP and VP jointed precoding algorithm is proposed. In order to optimize the performance of these two algorithms, this paper discusses user selecting or sorting algorithms of them, namely, determining the VP precoding users according to CSI. Simultaneously, BER and capacity of these two algorithms are theoretically analyzed. The simulation results prove that the algorithm proposed in this thesis can obtain a certain BER performance gain. Finally, the proposed algorithm is simulated in the LTE link simulation platform. Results show that it can btained a similar performance gains.