Performance of a maximum-likelihood decoding metric for convolutional coded FH/BFSK systems over AWGN channels under partial-band jamming

Abstract

To provide an upper bound on the performance of the existing decoding metrics for convolutional coded frequency-hopped binary frequency-shift-keying (FH/BFSK) spread-spectrum communication system, an optimum maximum-likelihood (ML) metric is derived operating over additive white Gaussian noise (AWGN) channels under partial-band jamming (PBJ). It is shown that the proposed ML metric requires the side information of the signal energy, the noise-plus-jamming power spectral density and the computation of logarithm Marcum-Q function to implement. Moreover, a simplified suboptimum ML metric is derived, and its union bound on bit error rate (BER) serves as a tight upper bound on BER for the optimum ML metric. A lower BER bound is also obtained for the ML metric. Tightness of these bounds is validated by simulations. Performance comparison under the worst-case PBJ shows that the proposed ML metric outperforms various existing metrics over AWGN channels.