认知MIMO干扰网络中基于子空间跟踪的低复杂度干扰对齐算法

Low Complexity Interference Alignment Algorithm Based on Subspace Tracking in Cognitive MIMO Interference Network

认知MIMO干扰网络中,多对认知用户通信而相互干扰,传统算法采用干扰对齐获得最佳传输速率,却因涉及对高维矩阵的多次特征值分解,处理复杂过高而难以实际应用。为有效降低算法复杂度,提出一种基于子空间跟踪的低复杂度干扰对齐算法。该算法以干扰信号到期望干扰空间的投影距离最小化为优化目标,通过矩阵瞬时共轭梯度只对低维的编码矩阵和干扰抑制矩阵递归更新,避免了冗余、重复计算。仿真结果表明,所提算法有较好的收敛性能,与传统算法相比,线性运算复杂度由O(n3)降低到O(n2),且避免了非线性运算。

In cognitive MIMO interference network, interference exists due to communication among multiple pairs of users. Traditionally interference alignment is used to obtain the optimal transmitting rate, but it’s too complicated to be applied because of high dimensional matrix decomposition. A novel low complexity interference alignment algorithm based on sub space tracking is proposed. This algorithm aims at obtaining the minimum projection distance from interference signal to expected interference space, updating only low dimension encoding matrix and interference suppression matrix recursively by matrix conjugate gradient, so that redundant and reduplicated calculation is avoided. Simulation results indicate that the proposed algorithm has better convergence performance, its computation complexity has been reduced from O(n 3) to O(n 2), and nonlinear computation is avoided.