Low Probability of Intercept Communication Based on Structured Radio Beams Using Machine Learning

Jialiang Zhou,Shilie Zheng,Xianbin Yu,Xiaofeng Jin,Xianmin Zhang
DOI: https://doi.org/10.1109/access.2019.2955509
IF: 3.9
2019-01-01
IEEE Access
Abstract:A low probability of intercept (LPI) communication system based on structured radio beams using machine learning is proposed. Information symbols can be physically encrypted on the structured radio beam realized by superposition of plane spiral orbital angular momentum (PS-OAM) modes. Performance of decryptor based on support vector machine (SVM) is analyzed. The accuracy achieves higher than 99.5, which is much better than the interceptor based on k-means. The influence of the deflection angle, number of receiver array elements and the time delay of interceptor are further analyzed. Simulation results shows that the LPI scheme can achieve security communication within practical SNR range.
What problem does this paper attempt to address?