Adaptive decoding methods for block codes in frequency-hopped spread-spectrum communications

Abstract

It is well known that some form of forward error correction is required to obtain acceptable performance in frequency-hopped spread-spectrum (FH/SS) communications. For most error and erasure decoding schemes, erasures are preferable to errors because more erasures than errors can be corrected. Thus, the use of error-erasure decoding for block codes in FH/SS communications can improve the performance of the decoder substantially if an effective and reliable method is employed to decide which symbols should be erased. In this thesis, several attempts are made to improve the error-erasure decoding performance of block coded FH/SS communications in a rayleigh fading channel. First, we propose the adaptive decoding scheme for concatenated codes in FH/SS communications in the presence of a pulse-burst jamming and analyzed its performance. Concatenated coding schemes employing binary BCH inner code and Reed-Solomon (RS) outer code are investigated and the use of side information is allowed to decode both erasures and errors. The proposed scheme makes the decoder enable to adapt to the jamming level by switching between two decoding modes such that the decoded bit error probability can be reduced. The optimal threshold value for switching in this proposed scheme is derived. It has been shown that the proposed decoding scheme yields a significant lower bit error probability over a conventional decoding scheme. In addition, performance analysis and its variation of conventional and proposed decoding scheme in the case of imperfect side information are also presented. Second, we propose the effective error and erasure decoding scheme for frequency-hopped multiple-access (FH/MA) communications with M -ary FSK signaling and analyzed its performance in synchronous and asynchronous frequency hopping systems, respectively. RS code is used to correct erasures and errors caused by other-user interference. The performance measure that we consider is the average number of succ...