摘要： 传统确定性半盲算法在最优权值选择上效率较低。为此，提出一种基于噪声子空间的半盲方法。利用噪声子空间与信号子空间的正交关系，构建信道响应与噪声矢量间的约束，根据参考符号与对应接收信号间的卷积关系建立额外的约束，由最小二乘方法求解信道冲激响应。仿真实验验证了该算法的有效性及参考符号个数下限的正确性。

关键词: 半盲信道辨识, 单输入多输出, 二阶统计, 噪声子空间, 参考符号, 迫零均衡器

Abstract: The traditional deterministic semi-blind algorithm has low efficiency when finding the weighting parameter. This paper focuses on the deterministic semi-blind methods, and proposes a new semi-blind method based on noise subspace. Based on the classical subspace decomposition, an orthogonality property between the signal subspace and the noise subspace is exploited to build a linear system of equations; then, additional equations are derived by the referenced symbols; at last, the estimated channel is derived by least-square method. Simulation examples demonstrate the performance of the algorithm.

Key words: semi-blind channel identification, Single-input Multiple-output(SIMO), Second-order Statistics(SOC), noise subspace, Reference Symbols(RS), zero forcing equalizer

中图分类号: 

• TP312