A block floating point treatment to finite precision realization of the adaptive decision feedback equalizer

Abstract

A scheme for efficient realization of the adaptive decision feedback equalizer (ADFE) is presented using the block floating point (BFP) data format which enables the ADFE to process rapidly varying data over a wide dynamic range, at a fixed point like complexity. The proposed scheme adopts appropriate BFP format for the data as well as the filter weights and works out separate update relations for the filter weight mantissas and exponents. Overflows at the feed forward and the feedback filter output are prevented by certain dynamic scaling of the respective input, while overflow in weight update calculations is avoided by imposing certain upper bound on the algorithm step size μ which is shown to be less than the convergence bound. The proposed scheme deploys mostly simple fixed point operations and is shown to achieve considerable computational gain over its floating point based counterpart.