非线性有限冲激响应Volterra信道的盲辨识

Blind Identification of Nonlinear Finite Impulse Response Volterra Channels

提出了单输入多输出有限冲激响应 (SIMOFIR)Volterra系统的盲辨识及盲均衡的子空间方法 .首先描述了SIMO非线性Volterra信道模型 ,并将此模型转化为多输入多输出 (MIMO)信道模型 ;接着讨论了FIRVolterra系统盲可辨识条件 ,并给出了SIMOVolterra系统的盲辨识子空间算法 ;进一步讨论了VolterraFIR非线性信道的确定性盲均衡问题 ;最后用仿真实例验证了此算法的有效性 .所提方法与确定性的线性多信道方法相比 ,它的最大优点就在于所需约束条件少 ,此方法仅需要输入信号的自相关矩阵非奇异 ,且在低信噪比情况下 (SNR为 5dB)也有效 .

A subspace approach to blind identification and equalization of a nonlinear single-input multiple-output (SIMO) finite impulse response (FIR) Volterra system is proposed. Firstly, a SIMO nonlinear Volterra channel model is described, and is transformed into a MIMO blind channel model. Then, the blind identificable conditions for FIR Volterra system are discussed and a subspace approach to blind identification of SIMO Volterra system is given; furthermore, the deterministic blind equalization of Volterra FIR nonlinear channel is studied. Finally, the validity of this method is verified by a simulation example. The advantage of this method is that it requires less restriction conditions, i.e., it only requires that the auto-correlation matrix of an input signal is non-singular and can work well in the case of low SNR in comparison with the deterministic approach.