Coherent sequence detection of MSK signals under impulsive noise

Impulsive noise modeled by symmetric α-stable (SαS) distribution can be found in many communication scenarios, such as atmospheric noises in very low frequency and low frequency (VLF/LF) communication systems and network interferences in wireless communication networks, etc. Under these cases, as Gaussian noise based signal processing methods always have poor performance, it is necessary to design robust signal processing algorithms to combat with the impulsive noise. In this work, we focus on the coherent signal detection of minimum shift keying (MSK) under SαS noise, because MSK is widely used in VLF/LF communication systems. Based on the received phase-coherent pass-band MSK signal model, a sequence detection algorithm is proposed by using the Viterbi algorithm. Under SαS noise, as the maximum likelihood (ML) based branch metric in the Viterbi algorithm has no closed form and is hard to implement, a robust branch metric is first proposed based on a closed form approximation to the ML based branch metric. Furthermore, the symbol error rate (SER) performance of the proposed sequence detection algorithm is analyzed. Our analytical results match the simulation results well, and both of them validate the robustness of our proposed algorithm.