Research On Interference Alignment Technology In The Downlink Of Cellular Networks

The inter-user interference(IUI)and inter-cell interference(ICI)greatly restricted the spectrum efficiency exist in multi-cell multiuser downlinks.Hence,it is important to design effective interference management methods.Interference alignment technique is an effective interference management mechanism that can overcome the shortcomings of traditional interference management schemes,eliminate the interference completely,and increase system capacity significantly.With the increase of the number of cells and users,the more transmit and receive antennas and the amount of channel state information are required for interference alignment,which result in the great increase of the complexity.Hence,it is difficult to apply the interference alignment to the cellular networks,which has a large number of cells.The interference alignment algorithm based on the interference graph proposed in this thesis can not only eliminate the IUI and ICI,but also can improve system performance,which is independent of network size.This thesis mainly focuses on the following aspects:1.The average(per cell)degrees of freedom in the downlink of cellular based on the interference graph is analyzed.Take the interference alignment feasibility condition in the downlink broadcast channel into consideration,we aim to design an optimization method to get the maximum average(per cell)degrees of freedom.The number of variables in the constraints will increase with the increase of the number of users,which leads to the optimization problem be a nonlinear and discrete problem.To solve this problem,the topology of the interference graph is used to transform the optimization problem into a linear problem,which has a small number of variables.Finally,the expression of the average(per cell)degrees of freedom in the downlink of cellular is obtained,which provides an important reference for the design of the interference alignment algorithm in the scenario provided in this thesis.2.A grouping interference alignment algorithm based on interference graph is proposed.Firstly,the cellular model is established by the interference graph.Then,we choose several cells as a cluster and eliminate the interference in this cluster by grouping interference alignment algorithm.Finally,the interference among clusters is eliminated by clusters' translation and all the interference of the active cells in the cellular networks is aligned.The simulation results have shown that the number of transmitting and receiving antennas and the computational complexity have been reduced by the proposed algorithm,which leads to the improvement of the system performance with no limitation of the network size.3.The received information should be fed back to the transmitter by the proposed the grouping interference alignment algorithm,which increases the overhead of the system.In order to solve this problem,a cascaded precoding interference alignment algorithm is proposed.Firstly,we design a cascaded precoding matrix.The first precoder is obtained by SVD algorithm to eliminate the IUI in a cell and the second precoder is used to align the ICI.Compared with the grouping algorithm,the proposed algorithm can achieve theoretical freedom without any feedback.4.Based on the above works,a cascaded precoding interference alignment algorithm based on interference graph is proposed.After the clustering method is designed,the proposed algorithm can not only be used to eliminate all the ICI and IUI in the clusters,but also the interference among clusters.Simulation results has shown that the proposed algorithm can improve the system performance with no limitation of the network size too.