Adaptive Equalization: Theory and Practice
All communications systems are subject to channel distortions, at the least they are annoying at the most they are disastrous. The distortions can be modeled by an equivalent channel transfer function. The remedy embodied by equalization is the synthesis of a receive filter that removes the distortion introduced by channel.
Adaptive equalization determines the corrective filter in a dynamic manner based on the current channel transfer function. The same basic adaptive equalization principles (identification and correction) apply to both analog and digital communication systems. A model of the channel transfer function is determined based on information obtained from the transmitted signal, then a receive filter that mitigates the channel distortion is synthesized. Many methods exist for both the identification and correction processes of adaptive equalization.
This tutorial is divided into two sections: a theoretical review of adaptive signal processing and adaptive equalization and a practical presentation of adaptive equalization techniques in both an analog and digital communication system. The theoretical section includes a discussion of the adaptive linear combiner, the performance surface, gradient estimation, LMS and RLS algorithms and filter structures.
The practical section begins with analysis and simulation results, including fixed point considerations, for an NTSC ghost canceling system, covering the aspects of equalization for an analog/continuous signal. The practical section concludes with an investigation of the performance of the many types of adaptive equalizers utilized for digital/discrete signals: synchronously spaced, fractionally spaced, decision feedback, reference directed, decision directed and blind.
Please disable any pop-up blockers for proper viewing of this Whitepaper.