A Concurrent Algorithm for Blind Adaptation of DFE

Introduction: DFEs have been widely used to remove intersymbol interference from the data received through communication channels [1]. In situations where blind adaptation of DFEs is required, algorithms based on the constant modulus criterion [2] for joint updating of the feedforward and feedback filters may converge to degenerated solutions. This occurs when the signal at the equalizer output is independent of its input [3]. To avoid this problem, Szczecinski and Gei [3] proposed a new criterion based on the constant modulus cost function with constraint on the feedforward and feedback filters, which is minimized by the stochastic DFE-CMA-FB (DFE Constant Modulus Algorithm with Feedback). In the context of blind adaptation of Linear Transversal Equalizers, based on [4], Chen [5] proposed to operate a Soft Decision-Directed (SDD) algorithm concurrently with CMA for M -QAM signaling. At the cost of a moderate increase in computational complexity, CMA+SDD presents an improvement in equalization performance over CMA. Motivated by the good behavior of CMA+SDD, we propose a concurrent algorithm for blind adaptation of DFEs. It is named NDEG-SDD-CMA (Non-Degenerative SDD-CMA) and is a concurrent combination of DFE-CMA-FB and the SDD algorithm for M -QAM signaling.