The Research On Interference Alignment For Multiusers MIMO Network Via Max-min SINR

Multiusers interference is the difficult problem in multiusers Multiple Input Multiple Output wireless network. Interference alignment(IA) is a novel method for interference suppression. While there exist the phenomenon of unfairness of resources, the quality of service of multiusers is poor, which even can’t meet the requirements of communication. We will apply the Maximize the Minimum Signal to Interference plus Noise Ratio(Max-Min SINR) strategy to solve this problem. Firstly, we study interference alignment, then establish the Max-Min SINR optimization model for IA to improve the unfairness of the allocation of resources between multiusers. The design of precoding and interference suppression matrix is explained, and the power allocation is analysed, which provide the basis of Max-Min SINR optimization for IA.The multiusers MIMO symmetrical network is studied with perfect channel state information. We analyse the typical SINR optimization IA algorithm in MIMO symmetrical network, include MLI algorithm and ECCAA. MLI algorithm is a typical interference leakage suppression method for IA, while ECCAA is a representative Max-Min SINR optimized means for IA. A joint optimization combining power allocation for IA using Max-Min SINR is proposed to improve the unfairness of the allocation of resources between multiusers. A utility function is also presented to evaluate the fairness of resources allocation every users. The simulations are conducted and indicate that the joint optimization combining power allocation for IA using Max-Min SINR can not only increase the minimum SINR of the system, but also obtain fast convergence rate and receive highest utility ratio, at the cost of sum rate.Then, we study the multiusers MIMO asymmetrical network which is get close to the practical communication system. The clustering for IA is discussed to solve the interference suppression problem when the number of users exceeds the feasibility of IA. The soft cluster algorithm is provided and next combined with the modified MLI. According with the game theory, strict potential game model is raised, and a method for IA based on the Max-Min SINR game rules is proposed in MIMO asymmetrical network. Simulations is presented and show that the IA method based on the Max-Min SINR game rules can better improve the minimum SINR of the MIMO system, and achieve higher sum rate in comparison with the combination of soft cluster and modified MLI. When the number of users approximate the feasibility of IA after clustering, the utility ratio of IA method based on the Max-Min SINR game rules is lower.