Improvements in the channel Equalizer Performance Using Modified LMS and BP Algorithms
This paper introduces a comparative study in the communication channel equalization problem. Different types of linear and nonlinear channel models including linear and nonlinear phase are considered in this study. Adaptive linear filter and neural networks are used to imitate different equalizer models. The equalizer models are tested using different transmitted signals with different characteristics. A modification in the learning algorithm driving each model is proposed to obtain the minimum mean-squared error in the recovery process of transmitted signals. The modified algorithms have demonstrated their effectiveness compared to other conventional techniques especially in the noisy environment.
Keywords: Adaptive fiters, Neural Networks, Channel Equalization.
Download Full-Text
ABOUT THE AUTHOR
Ashraf A.M. Khalaf
Ashraf.A.M.Khalaf has got the PhD degree in computer science and engineering from Kanazawa University, Kanazawa, Japan at 2000, and B.Sc. degree in electrical engineering from ElMinia University, ElMinia Egypt at 1994. His research ineterst in adaptive filters, neural networks and signal processing applications in the field of electrical communication engineering. He is a member in IEEE and related societies since 12 years and he is working in Electrical engineering Dept., ElMinia University, Egypt
Ashraf A.M. Khalaf
Ashraf.A.M.Khalaf has got the PhD degree in computer science and engineering from Kanazawa University, Kanazawa, Japan at 2000, and B.Sc. degree in electrical engineering from ElMinia University, ElMinia Egypt at 1994. His research ineterst in adaptive filters, neural networks and signal processing applications in the field of electrical communication engineering. He is a member in IEEE and related societies since 12 years and he is working in Electrical engineering Dept., ElMinia University, Egypt
