Low-Complexity Soft and Hard- Binary Bose- Chaudhuri-Hocquenghem (BCH) decoding for IEEE 802.15.6 Machine-To-Machine Communications

Abstract:
IEEE 802.15.6 Machine-to-Machine (M2M)1 applications require reliable communications at lowenergy consumption. This fact makes BCH codes an interesting option for IEEE 802.15.6 communications since their algebraic nature permit them to obtain the positions of the erroneous bits with mathematical equations of reduced complexity. Moreover, if the decoder is designed with knowledge about i) Galois Field properties ii) Frame Structure and iii) Cross-layer data, then it is possible to speed up all the stages BCH’s decoder. Thus the obtained hard-BCH decoder is very computational efficient, which permits to implement practical soft-BCH decoders. Additionally, this work also compares the strategies for estimating the true WER based on Syndromes and Cyclic Redundancy Code (CRC). The results show that it is possible to reduce the processing time of the Chien’s search till two decimal scales which permit to obtain practical soft-BCH decoders. The estimation of the true WER using CRCs is much better than the Syndromebased whereas the use of both strategies obtains a WER quite close to the true one.