Interference suppression for fast frequency-hopped communication systems over fading channels

Abstract

In this thesis, we consider interference suppression for fast frequency-hopped (FFH) communication systems over fading channels. Three types of interference, namely, partial-band noise jamming (PBNJ), multitone jamming (MTJ) and multiple-access interference (MAI), are of interest. To suppress the effect of PBNJ, we propose a maximum-likelihood (ML) receiver for frequency-selective Rayleigh-fading channels and derive the corresponding analytical bit-error rate (BER) expressions. We then focus on a more general frequency-selective Rician-fading channel, in which an optimum ML receiver structure is proposed. In addition, two suboptimum ML receivers and their corresponding analytical BER expressions are presented. Furthermore, the linear-combining and product-combining receivers are also investigated under the same channel conditions. To mitigate the composite effect of MTJ and PBNJ, we present analytical BER expressions for an ML receiver under frequency non-selective Rayleigh-fading channel conditions. Following that, frequency-selective Rayleigh-fading channels are considered. System performance of the linear-combining, product-combining and ML receivers is studied and compared. We then investigate the effects of timing and frequency offsets on system performance under frequency non-selective Rician-fading channel conditions. Numerical results show that the FFH system is very sensitive to the timing and frequency offsets. Lastly, we focus on the suppression of MAI. An optimum ML receiver and two suboptimum ML receivers, namely, the ML-A and ML-B receivers, are proposed for synchronous systems over frequency non-selective Rician-fading channels. Analytical BER expressions of these two proposed suboptimum ML receivers are presented. In addition, we study the system performance of the linear-combining, product-combining and clipper receivers, and compare their BER performance with the proposed optimum and two suboptimum ML receivers. Performance comparisons show that the proposed optimum ML receiver has the best BER performance, and the ML-A receiver outperforms the other four receivers.