An Efficient BCH Decoder for WBAN Applications

Abstract

In this work, a Bose-Chaudhuri-Hocquenghem (BCH) code decoder based on the Meggitt algorithm is implemented for wireless body area network (WBAN) applications. Compared to other implementations based on the popular Peterson algorithm, the proposed decoder has the extra capability of checking the syndrome for the decoding result, and is able to detect 20 times more failed decoding results without introducing additional complexity. When the proposed decoder is used as the hard-decision decoder (HDD) core of a Chase-2 decoder, the bit-error rate (BER) performance can be improved by about 0.3 dB because of the reduction in undetectable errors. Based on a comparison of the indices for corrected bits and those of the flipping bits, a test pattern reduction technique is proposed for the Chase-2 decoder that does not sacrifice the error-rate performance. The synthesis results show that the hardware complexity for the HDD core can be reduced by more than 50% using a 90 nm standard cell technology.