A new approach for efficient SISO equalizer with modified least mean square algorithm

In the presented paper, VLSI Architecture for linear turbo equalization are studied for constituent soft-input soft-output equalizer blocks. Linear Turbo Equalizer is the joint form of equalization and decoding in which soft information is iteratively exchanged between the soft-input soft-output equalizer and soft-input soft-output decoder. Linear turbo equalizer exhibits better performance in comparison to Decision Feedback Equalizer in terms of Bit Error Rate (BER) for Proakis channels. In this paper, modified Least Mean Square adaptive algorithm is selected for implementation of soft-input soft-output equalizer as hardware intricacy required to design equalizer is small in comparison to other equalizer. The modified Least mean square adaptive algorithm separate reckons of error-computation block and weight-update block. It also allows performance of optimal pipelining by using feed-forward filter retiming of both these sections distinctly to minimize the number of pipeline stage and adaptation delay.