VLSI Design for SISO Equalizer using Different Algorithm

Abstract

This thesis addresses the imperative of low power consumption, high processing speed, and reduced area in wireless and mobile communication systems. The first segment focuses on developing an adaptive decision feedback equalizer (DFE) with a novel algorithm aimed at minimizing power consumption and computational complexity while maximizing speed. By implementing a sign-normalized block-based least mean square (LMS) approach, the algorithm significantly reduces power requirements and processing time without compromising performance. Partitioning incoming data into non-overlapping blocks and conducting frequency domain filtering operations not only enhances speed but also contributes to area reduction by streamlining processing tasks. Moreover, the normalization of tap weight vector correction by the tap input vector enhances power efficiency and facilitates convergence in the mean squared sense, thereby reducing power consumption while maintaining performance standards. Simulation studies confirm the superior power efficiency, processing speed, and area reduction achieved by the proposed algorithm compared to conventional methods, making it an ideal choice for resourceconstrained wireless and mobile communication systems. In the second part, the focus shifts to linear turbo equalization, where the emphasis lies on achieving high-speed processing and reduced area while maintaining low power consumption. By integrating equalization and decoding functions and employing the LMS adaptive algorithm for equalization and the sliding window log-MAP algorithm for decoding, the proposed approach achieves significant improvements in processing speed and area reduction without sacrificing BER performance. Leveraging fixed-point representation to approximate the Log-MAP decoding algorithm strikes a balance between performance and complexity, further enhancing efficiency. Additionally, the selection of the LMS algorithm for equalization minimizes hardware complexity, contributing to area reduction and power efficiency. The resul