A receiver-based variable-size-burst equalization strategy for spectrally efficient wireless communications

Abstract

Strategies for improving spectral efficiency in wireless communications, while maintaining quality of service, are proposed. The idea is to make the communication system adapt to the encountered environment. To this end, a time-varying channel model explicitly characterizing changing environments is developed. Under this model, methods for tracking channel stationarity are proposed based on analysis of variance and minimum description length. Tracking enables efficient construction of variable-size data bursts for improved equalization at the receiver. The performance is evaluated first in a training-based equalization framework, then extended to a semi-blind situation. In either case, the obtained results demonstrate feasibility of variable-size burst construction for improving spectral efficiency. The proposed methods are applicable to burst-by-burst or block-based equalization systems, without the requirement of feedback.