A Set-Membership-Based Blind Equalization Algorithm for High-order QAM Signals

This paper presents a new blind equalization algorithm based on the concept of set-membership filtering (SMF). We first employ a new strategy in the stop-and-go algorithm (SGA) to improve the update reliability by substituting the Sato-like error with multimodulus error and involving a penalty function. Then an adaptive algorithm is derived in terms of the SGA and the set-membership binormalized data-reusing LMS algorithm. During the coefficients update, the proposed algorithm uses two constraint sets to construct reliable solution space. Simulation results demonstrate that compared with other algorithms including SGA, the proposed algorithm provides superior performance with regard to both the convergence rate and the steady-state mean square error (MSE), and is comparatively efficient for high-order QAM signals