Adaptation of Butler-Cantoni method for fast start-up equalization

Abstract

Fast start‐up in adaptive equalization is indispensable in communication systems that require on‐line performance. In this article, a new type of adaptive linear FIR equalizer composed of two FIR filters and called the adaptive Butler–Cantoni (ABC) equalizer is proposed, and speed‐up of convergence in the training mode is attempted. The ABC equalizer consists of a channel estimator and an equalization filter. The channel estimator updates coefficients by the LMS algorithm. At the same time, the variance of the additive noise is computed. Based on these results, the coefficients of the equalization filter are successively derived by the Levinson–Trench algorithm, and the estimated value of the transmitted signal is computed. The computational complexity of the ABC equalizer is proportional to the square of the tap length but is not more than that of the RLS equalizer. By computer simulation, it is demonstrated that the ABC equalizer provides better results than the LMS and RLS equalizers. © 1999 Scripta Technica, Electron Comm Jpn Pt. 3, 82(10): 9–17, 1999

Coefficients--Delay Simultaneous Adaptation Scheme for Linear Equalization of Nonminimum Phase Channels

“…As shown in Sect. 3.3, the proposed delay adaptation method can be implemented with the simple procedure, which is combined with the conventional ABC equalizer [9]. In Step 2, the proposed scheme requires the average operation of (67), which is reexpressed as…”

“…Recently, Shimamura et al have proposed an adaptive LTE involving a channel estimator, which is called the adaptive Butler-Cantoni (ABC) equalizer [9], [10]. The ABC equalizer converges faster than the LMS and RLS transversal equalizers, because it is indirectly adapted from the results of the channel estimator by use of the Butler-Cantoni non-iterative method [11] for each iteration.…”