Abstract
Wireless communication systems are affected by
inter-symbol interference (ISI), co-channel interference in
the presence of additive white Gaussian noise. ISI is primarily
due to the distortion caused by frequency and time selectivity
of the fading channel and it causes performance degradation.
Equalization techniques are used to mitigate the effect of ISI
and noise for better demodulation. This paper presents a novel
technique for channel equalization. Here a Signed Regressor
adaptive algorithm based on FLANN (Functional Link Artificial
Neural Network) has been developed for nonlinear channel
equalization along with the analysis of MSE and BER. The
results are compared with the conventional adaptive LMS
algorithm based FLANN model. The Signed Regressor FLANN
shows better performance as compared to LMS based FLANN.
The equalizer presented shows considerable performance
compared to the other adaptive structure for both the linear
and non-linear models in terms of convergence rate, MSE
and BER over a wide range.
Users
Please
log in to take part in the discussion (add own reviews or comments).