基于FastICA的盲干扰对齐算法

为克服现有的干扰对齐方法中发送端信道传输信息不足的问题, 提出了一种新的盲干扰对齐（BIA: Blind Interference Alignment）算法。在K用户MIMO(Multiple Input Multiple Output)干扰信道中, 发射端已经建立起干扰对齐机制条件下, 依照盲源分离(BSS: Blind Source Separation)理论, 将IA（Interference Alignment)模型与BSS模型相结合, 在接收端采用FastICA(Fast Independent Component Analysis)算法, 建立以干扰对齐为模型的目标函数。对数据进行中心化, 通过逐个分离的方法, 单独分离每个独立分量, 从而提取期望信号。Matlab仿真结果表明, 该算法能在未知信道状态信息的情况下有效实现干扰对齐, 同时具有较好的误码率。

FastICA for Blind Interference Alignment Algorithm

To overcome the problem of the interference alignment schemes without CSIT(Channel State Information at Transmitters), we present a new IA(Interference Alignment) algorithm with blind source separation for K user MIMO(Multiple Input Multiple Output) interference channel on the basis of already built interference align
ment design at transmitters. According to the BSS(Blind Source Separation),we build the objective function for the model of the interference alignment. We
centralize data by the method of separation individually, separate out each independent component, so that the desired signal is extracted by FastICA(Fast
Independent Component Analysis) algorithm at receivers. Simulation results in Matlab show that this algorithm can effectively realize interference alignment w
ithout CSIT, and keep good BER（Bit Error Rate）performance.