An Interference Alignment Scheme Based On Degree Of Freedom Scheduling In MIMO Interference Networks

Nowadays, with the rapid development of wireless communication, the type of wireless services and the architecture of networks become more and more complicated. The higher pursuit for quality of service leads to more severe interference problem, which impacts the network performance seriously. Thus, finding an effective way to manage the interference is quite significant.Interference alignment (IA) is an emerging coding technique, which can make interference signals aligned and then zero-forced, and meanwhile leave non-interfering subspace for useful signals through coding matrices at transmitters and receivers. Compared with traditional approaches, IA can eliminate interferences completely, and reach more degree of freedom (DoF) and greater network throughput.To further improve the performance, considering multiple-input multiple-output (MIMO) interference network, based on IA scheme, a DoF allocation algorithm is proposed in this paper. The algorithm allocates DoFs among users according to the channel quality with max-sum-rate as the objective function and feasible conditions of IA as constraints, where users with high quality channels get more DoFs while those with poor quality channels get less, to improve the performance of the network.Moreover, transmitted power allocation among different data streams is also considered in this paper, and a power allocation scheme based on maximal rate is proposed. Through simulation results, the improvement brought by the schemes can be seen obviously.Besides, considering that in practical systems, the channel estimations may have errors, which may cause wrong results for DoF allocation, two algorithms considering imperfect channel state information are presented. They are based on exhaustive search and discrete stochastic optimization respectively, and can achieve the optimal allocation scheme with best average performance. Furthermore, considering the requirements of users, a user demands based DoF allocation algorithm is also proposed, which is combined with the max-sum-rate based DoF allocation algorithm to balance the network performance and user fairness.