Research Of Interference Alignment In Wireless Communication System

Along with the increasing requirement of the transmission rate,the contradiction between limited spectrum resources and higher data transmission rate of the becomes very prominent.Therefore, the next generation of wireless communication system adopt multi-antenna technology, multi-antenna technology not only achieves higher spectralefficiency, but also reduces the error rate. However, interference of multi-antenna technology has become the primary factor of restricting the capacity of the system, especially in the multi-user MIMO system, if there is no effective Interference Management Scheme, the interference will reduce or counteract the gain brought by the high spectral efficiency. In this context, the interference alignment technology emerges naturally, it overlaps interference signal at the receivers,compresses the space occupied by the interference signal. The research results show that using interference alignment technology can make the system capacity increase linearly with the increasement of antennas.This thesis shows different interference channels and illustrates them.In terms of sum-rate performance,the Minimum Interference Leakage algorithm is sub-optimal, therefore we propose a distributed interference alignment algorithm which plays a role of adjusting interference from unintended transmitters via different weight w.Then, we will further reduce the interference through "rotation" operation.The simulations show that the sum-rate of proposed alignment outperforms the existing minimum leakage algorithms especially at low SNR region(up by 30%) and the interference existing in the system depends on the weights.Finally, based on the weighted optimization proposed in this thesis, "rotation" could further improve the system performance of sum-rate.In order to avoid vast information exchanged between the transmitter and receiver.our algorithms proposed in Chapter Four runs at transmitters only, The first algorithm not only reduces the interference subspace chordal distance,but also makes the interference subspace away from the useful signal subspace.Based on the chordal distance of subspace, we build the target expression and then get the local or global optimal solutions along the steepest descent direction of the target expression.The second algorithm reforms the optimization problem on the Stiefel manifold.On the manifold,we can search to find the local or global optimal solutions.Finally,the simulation results show that both algorithms could achieve effective interference suppression and enjoy better performance of sum-rate.In terms of convergence,the second algorithm is better.