Modern Signal Processing Research On Anti-Interference Of Spread Spectrum Communication

With the superior concealment performance, implementation convenience of network access and other advantages, spread spectrum communications become the world wide mainstream technology of anti-jamming. However, by solely relying on the spread spectrum technology, it is very difficult to meet the future anti-jamming needs of the military battlefield communications under complex electromagnetic environment. There are strong needs for new technology to improve anti-jamming capability of the spread spectrum communication system. This dissertation focuses on the development of modern signal processing methods and technology for anti-jamming in spread spectrum communication.We first explore the possibility of utilizing blind separation technology to defend against jamming in spread spectrum communication. Then we propose blind separation anti-jamming system architecture for P2P and networking spread spectrum communication. And we design a new negentropy aided anti-jamming algorithm based on blind separation for DSSS system. Under strong noise environment, average filter is adopted in the refinement of DSSS blind separation anti-jamming algorithm to take full advantage of the actual high sampling rate. The composition of user asynchronous delay is analyzed. Its influence on the convergence of the blind separation anti-jamming algorithm has been investigated and verified by triangle inequality.The blind separation frequency-hopping anti-jamming algorithm is proposed to avoid the adverse effect of the frequency transition. The algorithm’s anti-jamming ability is verified under related interference and non-related interference by MATLAB simulation and real hardware experiments. As we observe, the time-frequency distributions of a frequency-hopping signal and most kinds of jamming are usually quite different. Accordingly, we propose a new blind separation anti-jamming algorithm based on joint time-frequency distribution which separates the signals by matrix joint diagonalization. With the known frequency-hopping graph, a semi-blind separation anti-jamming algorithm is also designed. We design a nonlinear transformation and the corresponding detection method dedicated for chirp interference, which can produce chirp signal compress sensing dictionary with excellent energy aggregation characteristic in a self-learning way. From the inconsistent time-frequency distributions between multi-tone interferences and the frequency-hopping signal, their corresponding self-learning compress sensing dictionaries have been constructed to anti strong interference under low signal-to-noise ratio condition. Experiments show the self-learning dictionaries can help greatly for the de-modulation procedure in real-time under challenging conditions.We propose a novel polymorphic self-learning compress sensing anti-jamming algorithm for frequency-hopping communication (SLCSFH) by the combining self-learning dictionary construction and compress sensing technology. Excellent anti-jamming ability of the proposed SLCSFH algorithm is verified under low signal-to-noise ratio and strong interference condition by MATLAB experiments. We also propose a single channel anti-jamming algorithm based on SLCSFH which removes the interference component detected during the self-learning procedure.Finally, we show that internal pattern is a widely established inherent characteristic of signals. By using the self-learning method, we can get special and tiny sparse domains which are more appropriate for the destination signal. Then we can reconstruct the signal efficiently by using the compress sensing technology. The proposed self-learning reconstruction method gives a pragmatic direction for compress sensing technology. We give detailed illustration of the proposed method by an excellent application in image super resolution reconstruction.