Myriad MSK Signal Detection for VLF/LF Communication.

Guosheng Yang,Jun Wang,Guangrong Yue,Shaoqian Li

DOI: https://doi.org/10.1109/milcom.2016.7795401

2016-01-01

Abstract:In VLF/LF communication systems, atmospheric noise is the dominant interference, which is usually highly impulsive and can be effectively modeled as symmetric α-stable (SαS) distribution. It has been known that communication techniques designed under Gaussian noise assumption always have poor performance under impulsive noise. In this work, we focus on the detection problem of minimum shift keying (MSK) signals under SαS noise, because this modulation is widely used in VLF/LF communication systems. As SαS distribution usually has no closed form probability density function (PDF), the optimal maximum likelihood (ML) based detection of MSK signals is hard to implement. To address this problem, we proposed a closed form approximation of the branch metric used in the Viterbi algorithm (VA) for MSK coherent detection, and provided a closed form approximation of the local optimal MSK non-coherent detection. Compared with ML based algorithms, the proposed algorithms have closed form and are simple to implement. Simulation results show that the bit error ratio (BER) performance of the proposed MSK coherent detector can closely approach that of the ML based coherent detector. Meanwhile, the proposed MSK non-coherent detector can attain the similar BER performance as that of the ML based non-coherent detector in low signal-to-noise ratio region.