一种改进的稀疏多径信道盲辨识算法

摘要

针对无线多径稀疏信道,利用有效近似思想,提出了一种改进的基于矩阵外积分解的信道盲辨识算法.算法首先采用改进的VIA准则精确估计稀疏信道“有效部分”的阶数；然后使用矩阵外积分解算法估计信道冲激响应的“有效部分”,为了降低噪声及信道冲激响应中“零抽头”部分的影响,本文提出一种新的噪声方差估计方法,利用重新构造的自协方差矩阵,能够得到比较精确的噪声方差估计值,提高了外积分解算法在中、低信噪比条件下的盲辨识性能；最后利用盲辨识结果进行反卷积,恢复出发送信号.与现有算法相比,本算法不仅降低了对信噪比的要求,而且克服了子空间算法的相位偏转问题.仿真实验以及对SPIB微波信道测试结果验证了本文算法的有效性.%Based on the idea of the effective approximation, this paper uses the matrix outer-product decomposition and proposes a new blind identification and blind equalization algorithm to solve SMC (the sparse multi-path channel) problem in the wireless communication. Firstly, an improved VIA principle is adopted to precisely estimate the order of the effective part of the SMC. Then the channel coefficient of the effective part is estimated with the improved matrix outer-product decomposition. In order to eliminate the interference of the noise and the ZT(zero-taps) in the channel impulse response, this paper adjusts the noise variance estimation method so that the performance of the blind identification is improved when the SNR( signal-to-noise ratio) is low. Finally, based on the results of the estimation, the received signal is deconvolved to derive the transmitted signal. Compared to the existing method, the algorithm proposed relaxes the SNR requirement of the estimation, and overcomes the phase distortion problem in Subspace Algorithm(SSA). The validity of the proposed method is verified via numerical simulations and the test results on SPIB microwave channel.