Blind Compensation of Non Linear Channels

Non-linear distortions related to the band-limited signals results in the spectral spreading of the original signals, The proposed technique is to suppress the unwanted spectral contents with the help of Adaptive polynomial filtering. It is most useful in the scenarios where a direct access to the source or undistorted signals is not possible,either due to the cost factor involved or is not feasible at all. The developed algorithm is based on power series filter with adaptive coefficients blindly identified with LS criterion, and the Discrete Cosine Transform is used to get the spectral knowledge of the received signal, and finally the energy of the unwanted spectral contents are minimized, most efficiently to zero. In this book, memory less non linearity has been addressed and the signals are either periodic narrow band signals or the band limited speech signals, so the people working in these fields would find this work useful and it also provides a basis to work on the non linear distortions with memory, and in that case the non linear functions would be modeled using the Volterra series.

Waheeb A Butt, MSc: BS Electronics Engineering, International Islamic University, Pakistan. MSc in Communication and Signal Processing with distinction from Newcastle University, UK. Looking forward to pursue a PHD in the communication technologies.