Channel Estimation using Conventional Methods and Deep Learning

Despite the evolution of wireless communications, there are some undesirable effects on the signal transmitted through a wireless channel resulting from its physical characteristics. Thus, signals arriving at the receiver are attenuated, distorted, and delayed. Channel estimation is an important part of receivers. A perfect channel estimation compensates for the effects of the channel and restores the original signals. This paper aims at implementing different channel estimation techniques like Least squares estimation, Minimum mean square estimation. Also equalization techniques like Adaptive equalizer - LMS, CMA equalizer and MLSE are analyzed. The implementation is done initially using MATLAB software and then using GNU Radio. Deep learning is also explored for channel estimation in Orthogonal Frequency Division Multiplexing in MATLAB.