Study On Signal Detection Technologies Of Large Scale MIMO Systems

Large scale MIMO has the advantages of high transmission efficiency, high spectral efficiency, high energy efficiency and high anti-interference ability, which has become a hot research topic in the field of wireless communication. Signal detection is one of the core technologies of large scale MIMO. It has an important influence on anti-interference ability and reliability of whole system.With the number of antennas increasing, the technology of conventional MIMO signal detection is applied to large scale MIMO, which has problems of high complexity or poor performance. In order to solve these two basic problems, the thesis focuses on two heuristic algorithm Likelihood Ascent Search(LAS) and RTS(Reactive Tabu Search). The main results are summarized as follows.(1) The large-scale MIMO system model is provided. The performance and complexity of the current signal detection algorithms is analyzed. The emphasis is laid on the LAS algorithm, RTS algorithm and the derivative algorithms of these two. With the experimental simulation, the effects of the starter performance, the starter number and the maximum searching times on the performance of BER and complexity is analyzed, providing an entry point for the researching work of this thesis.(2) LAS signal detection algorithm is easy to fall into the local minimum resulting in poor performance and its improved algorithms improve performance while they increase the complexity of signal detection, so the thesis proposes a new improved LAS algorithm based on genetic properties. The algorithm rooted in traditional LAS algorithm judge the quality of the solution vector by analyzing the statistics of ML cost function under error-free reception conditions. Then correct the number of iterations needed for the algorithm by the quality of the solution vector. At the same time inherite optimal solution vector, randomly change the value of optimal solution vector at certain positions and get the initial vector of the next iteration. The simulation results show that the proposed LAS algorithm combined with genetic properties further improves the performance of the algorithm and simplifies the complexity by reducing the number of LAS iterations.(3) In order to improve the perfomance of RTS algorithm under high order modulation, the thesis proposes a novel layered algorithm based on adaptive change of equivalent channel dimensions. The algorithm rooted in layerd algorithm adaptively changes the maximum number of searching steps according to the dimension of equivalent channel matrix. When the dimension of the equivalent channel matrix is small, the proposed algorithm avoids the circuitous search by small maximum number of search step.When the dimension of the equivalent channel matrix is large, the proposed algorithm makes a wide search range by large maximum number of search step,to enhance search capability and improve detection performance. The simulation results show that the proprosed algorithm improves the detection ability and the reliability of the system.