Blind Deconvolution for SISO FIR Channels Based on Autocorrelation Function

Shefeng Yan,Yuanliang Ma
DOI: https://doi.org/10.1109/icnnsp.2003.1281138
2003-01-01
Abstract:A new algorithm based on autocorrelation function of single input for blind deconvolution is proposed. We deduce a LMS iterative algorithm for blind deconvolution by minimizing the squared sum of received signal autocorrelation function over a region excluding the zero delay point. In this algorithm, only second order moment is adopted, which makes it widely applicable for Gaussian signals as well as non-Gaussian signals. Computer simulation results agree well to the theoretical analysis.
What problem does this paper attempt to address?