Maximum likelihood channel estimation based on nonlinear filter
Min Shen · Journal of Chongqing University of Posts and Telecommunications · 2008
For long finite channel impulse response,accurate maximum likelihood channel estimation is computationally high cost due to high dimension of parameter space,and approximate approaches are usually adopted.By utilizing the suppression of noise and extraction of signal of the nonlinear Teager-Kaiser filter,a likelihood ratio of channel estimation is defined to represent the probability distribution of channel parameters.Maximization of this likelihood function leads to initially searching the extrema of path delays and then the complex attenuation.Computer simulation is conducted and the results show performance improvements of joint detection as compared to the non-likelihood approach.