Research Of Interference Alignment Technology In Multi-cell MIMO Systems

With the rapid development of wireless communications, the number of users in communication systems increasingly becomes larger and communication environment is becoming more and more complex, so the existing spectrum resources become particularly valuable. How to improve the spectrum efficiency under limited spectrum resources to support higher rates, higher reliability and higher quality services for future wireless communication networks have become an important research topic.In order to improve spectrum efficiency, frequency reuse factor one is widely used in current cellular systems. In this case, interference from adjacent cells which use the same frequency band will become severely and this inter-cell interference have become a key factor constraining wireless communication. Interference alignment (IA) as a new coding technique to eliminate interference, has become a hot research topic because of the advantage of degree of freedom(DOF). It can suppress inter-cell interference efficiently and improve the cell-edge throughput.In this paper, we mainly studied the IA technology in multicell MIMO systems. Firstly, the classic IA algorithm is introduced when cell number K is3. Global channel state information (CSI) is required although the algorithm complexity is low and it can achieve the optimal DOF. Then four iterative IA algorithms which only need local CSI is introduced:Distributed IA Algorithm、Max-SINR IA Algorithm、 Alternative Minimization Algorithm、Joint Signal and Interference IA Algorithm. Simulation results show that the performance of distributed IA algorithm and alternative minimization algorithm is identical when the channel reciprocity is considered. Max-SINR IA algorithm performers better than distributed IA algorithm because the signal strength is considered. Joint signal and interference IA algorithm reduces both the interference power that "leaks" into the signal subspace and the signal power that "leaks" into the interference subspace, the tradeoff between the two parts is controlled by a weight parameterω. The simulation results also show that joint signal and interference IA algorithm give the same results with distributed IA algorithm if ω=0.In addition, we propose an IA algorithm based on antenna selection which antennas are selected by maximizing channel matrix norm. Each base station in multi-cell MIMO systems is equipped with more antennas than users. Firstly all base stations select out the antenna subset which has larger Frobenius norm. Then IA technology is used to eliminate inter-cell interference after antenna subset is selected. Simulation results reveal that the proposed algorithm outperforms the classic IA algorithm with no antenna selection and it performs better when more transmitting antennas are equipped at BS sides.