基于联合QR分解的干扰对齐算法

Interference Alignment Algorithm Based on Joint QR Decomposition

随着频率资源日益紧缺,全频率复用技术受到重视,而该场景下小区间干扰和小区内干扰是制约系统容量的重要因素。分布式迭代干扰对齐作为一种干扰管理技术,能够大大提高小区的系统容量,但是具有复杂度高的缺陷。针对3小区2用户的MIMO下行干扰信道模型,提出了一种能够对齐小区间干扰、小区内干扰的线性干扰对齐算法。该算法对联合信道矩阵进行QR分解,再依据各小区等效信道模型的特征,利用最小化干扰泄露以及迫零算法对齐小区间以及小区内干扰,每用户能够实现其信道空间维度1/3的自由度,而接收端只需进行一次预编码处理,大大降低了接收端的复杂度。仿真结果表明,相比于分布式迭代干扰对齐算法,在实现每用户相同自由度的情况下,能够在保证系统性能的同时大大降低复杂度,甚至在低自由度下性能更优。

With the increasing scarcity of spectrum resource, the full frequency reusing technique has been taken more seriously. However, the inter-cell and intra-cell interference become the bottleneck of the system capacity in that scenario. System capacity can be sharply im- proved with the interference management scheme named distributed iteration interference alignment, meanwhile, with the drawback of high complexity. For three cells and two users in each cell MIMO broadcast interference channel, a linear interference alignment algorithm is proposed that can align the inter-cell and intra-ceU interference. According to the characteristics of the equivalent channel model based on the QR decomposition of the joint channel matrix,each user can achieve Degree of Freedom （DoF） that is one-third of the channel space dimension with minimizing the interference leakage and zero-forcing. Moreover,just one pre-coding is processed in the receiver which can greatly reduce the complexity. The simulation results show that compared to the distributed iteration interference alignment al- gorithm with the same DoF for each user, the proposed algorithm can significantly decrease the complexity and guarantee system perform- ance, even better in lower DoF situation.