Sum-rate improved interference alignment in wireless MIMO interference networks

In this paper, three iterative algorithms (which are denoted by IA-Orth, IA-Power and IA-Rate algorithms in this paper) based on interference alignment (IA) are proposed to make further efforts to consider the useful signal subspace to improve the sum rate in the wireless multi-input multi-output(MIMO) interference networks. The proposed algorithms consider not only minimizing the spatial distance between different interference subspaces but also desired signal subspace from different aspects to further optimize the sum-rate performance. The optimization problems could be performed on the Grassmann manifold since the true domain of IA is over the Grassmann manifold. Moreover, considering different interference power levels at each receiver and IA feasibility constraints in practical environment, our previously proposed IA-Rate algorithm is modified to be adapted to a broader range of applications. It is on the basis of clustering and properly dealing with inter-cluster interference. A series of simulations show that our algorithms are superior to the conventional IA algorithms on the achievable sum rate and provide a very efficient way to realize IA in practical wireless MIMO interference channel environment.