Design of an Intelligent Demodulation Method for X-ray Spectrum Communication

Abstract X-ray communication based on X-ray spectrum modulation information is a novel communication method, which can effectively improve communication ability and anti-interference ability. In this paper, four characteristic X-rays are chosen to modulate information. The influence of channel attenuation, random noise in space, and miss-shooting rate caused by adjacent targets are discussed. Based on error-correcting output codes and a 10-fold cross-validation method, a neural network intelligent classifier is designed to realize intelligent information demodulation. Simulation results show that if the miss-shooting rate is lower than 0.4 and the photon retention rate after channel attenuation is higher than 0.0008, then the classifier can achieve a recognition success rate of more than 99%. This study provides reliable index requirements for the performance of the four-target material source equipment to be designed next. This paper also provides a theoretical basis for the subsequent further realization of space communication capabilities.