Research On The Application Of Interference Alignment In Device-To-Device Communication System

5G has become an important subject that people focus on recently. D2 D communication is a key technique in 5G, which causes interference while reusing cellular wireless resource. Interference alignment is exactly used for managing and reducing the interference in wireless communication. With the short communication distance, stable channel state, low mobility, and the channel reciprocity in TDD system, D2 D communication provides good basis for the application of interference alignment. Thus, this thesis mainly researches on the application of interference alignment in D2 D communication system, which improves the capacity and the resource utilization of mobile communication system.The application of interference alignment in SISO system is studied in this thesis. The interference alignment method in D2 D system without relay is firstly researched, with the simulation of its DOF upper-bound. Secondly, this thesis forms a virtual MIMO system with relay, applying a distributed interference alignment algorithm into D2 D communication system with relay, and then optimizes it. Simulation results show that the optimization scheme improves the system capacity and enhances the efficiency of the algorithm.Interference alignment in MIMO D2 D system is then taken into consideration. This thesis considers a system model with only D2 D users, then adds the interference that cellular users cause into the model, and proves the feasibility of alignment. A distributed interference alignment algorithm is researched in the system, and the proof of convergence is finally given. Simulation results show that the algorithm improves the system performance. Simulation of various numbers of antennas and iterations are also shown to analyze the system capacity and the complexity of the algorithm respectively.Grouping strategy based on the principle of interference alignment is researched lastly. By making Fuzzy C-means algorithm suitable for interference alignment, this thesis designs a position based algorithm for clustering and a channel based algorithm for grouping in each cluster, as well as a distance based grouping algorithm. Simulation results show the performance of each algorithm and indicate that the algorithms improve the system capacity.