Research On Interference Alignment In MIMO Multi-cell Networks

In a multi-cell MIMO communication system,the receivers at the edge of cell are strongly interfered by inter-cell and intra-cell signals resulting in a decrease of system capacity.As a novel idea for interference suppression in MIMO systems,the technique of interference alignment could reduce the spatial dimensions of the interfering signals and increase the degree of freedom of the system by designing the precoding matrix and the interference suppression matrix in the transmitter and receiver respectively and achieve a higher channel capacity than conventional methods.This thesis centers on the interference alignment methods in the uplink and downlink channels respectively in the MIMO systems with multi-cell multi-user and the specific contents are as follows:Aiming at the problem of ignoring the average spectral efficiency in the user scheduling for the multi-cell multi-user MIMO downlink channel,an opportunistic interference alignment algorithm based on the signal subspace is proposed in this thesis.Utilized the random orthogonal subspaces that base station broadcasts,this thesis firstly designs a coefficient that measures the degree of inter-cells interference,and the interference suppression matrix is conceived based on the inter-cell interference minimization criterion.Secondly,adopted the zero-forcing algorithm,the intra-cell interference is eliminated by the inner layer precoding.Finally,the base station sends data to the selected users with the smallest degree of interference in the downlink channel in between,the designed double-layer precoding method can process the intra-cell and inter-cell interference simultaneously.The simulation results manifest that,with the proposed method,the total capacity and energy efficiency of the system are increased largely and the system error rate is reduced.For the MIMO uplink multi-access channel,a limited feedback interference alignment algorithm based on upper triangular decomposition to deal with the situation with one cell-edge user in each cell is proposed.Differing from the existing limited feedback interference alignment algorithms that only the interference to be aligned being considered,the thesis offers a solution that triangularly decomposes the joint channels to obtain the equivalent model so that half of the existing interference is eliminated and the system sensitivity to the interference and the quantization errors of precoding is decreased and thus the stability of the system is enhanced.Then considering that different users get different interference,the joint channels is adjusted before the triangular decomposition for balancing the data rates of the users involved and the power is allocated.At last,the precoding matrix and the equivalent interference suppression matrix are designed by minimizing the interference leakage and minimizing the mean square error.The simulation results demonstrate that the algorithm proposed increases the system performance no matter there is feedback quantization error or not.