Rprop and improved Rprop+ based constant modulus type (RCMT) blind channel equalization algorithm for QAM signal

摘要

The aim of paper is to address the issues such as poor convergence, moderate BER, residual ISI associated with unsupervised (blind) weight adaptation technique employed for channel equalization. Two noble algorithms based on resilient back propagation framework have been proposed as solution for QAM signal. The widely accepted gradient descent based Constant Modulus algorithm suffers from poor convergence issues, high residual ISI and Moderate BER. Proposed solution assimilate Constant Modulus Algorithm (CMA) like error function into a general Rprop algorithm having two distinctive approaches of with/ without weight backtracking capability, taking into consideration of frequency selective nature of channels and additive gaussian noise. The algorithms exploit the advantage of Rprop mechanism, thus improving the convergence rate and have better ISI (inter symbol interference) suppression capability. Simulation of the proposed Rprop and improved Rprop+ based CM type (RCMT) algorithm proves their advantageously enhanced capabilities in terms of convergence, complexity, and residual ISI in comparison to the CMA and its variants. Proposed algorithm can also been used for M-PSK signal for which it shows similar performance.