Research On DSSS Signals Parameter Estimation Techniques In Electronic Counter Measure

Electronic warfare plays a very important battle role in suppressing enemy efficient communications in modern war. The military electronic counter measure, in the purpose of breaking enemy's electronic facilities and protecting our device in the work, intents for using electromagnetic and orient power to control electromagnetism. Under the condition of complicated unknown battle communications, it is necessary to detect, intercept and monitor enemy's communications signals and orient the incident signal source as well, and then measures with communication disturbances are applied to obstacle break and degrease the efficient enemy communication systems in operation, and ensure our connections are in work as well. So it is necessary to analyze the characteristics of the enemy sine signals and estimate the signal parameter obtained from the radio antennas. In the case of extremely low Signal-to-Noise Ratio (SNR), parameters estimation for the intercepted DSSS signals, DS sequence discrimination automatically and orient position realization in accuracy are extremely important problems to detect, disturb and anti-disturb the enemy communication systems within the electronic counter-measure.So far most signal feature extraction and analysis methods do not have sufficient robustness against non-stationary stochastic process. Signal parameter estimation in the case of low SNR is a very important research topic in the signal processing research area. This thesis aims to study the robust signal characteristics extraction methods with strong noise. The main innovative contributions of this thesis are as follows:Sample signal set is constructed with the different frequency feature vectors, signal feature vector set is obtained extracted from the samples with the proposed algorithms as well. Improved kernel principal component analysis algorithm lined with optimum multilayer classification structure is utilized to extract the signal feature, thus, estimated signal frequency is obtained. Simulation results show that the improved kernel principal component analysis algorithm can be applied to different frequency band and performs excellently on estimation.Conventional algorithms can not estimate the DSSS parameter under very low SNR, so forth order cumulant based adaptive method is proposed to increase the SNR of intercepted signals by multi-time cyclic iteration, and the algorithm also solves the problem of classical higher order theory with large computation. Simulation results show the algorithm is efficient for signal parameter estimation contaminated by stationary noise and non-stationary noise and the convergence of the algorithm is proved as well.Refer to the characteristics of strong autocorrelation for the pseudo random sequence and the way to estimate the pseudo random sequence by matrix decomposition, higher order based method is proposed to estimate the pseudo random sequence. Simulation results illustrate under low SNR, pseudo random sequence accurate synchronization and estimation can achieved hided in intercepted signals.In the one dimension array distribution, the different matrix between the forward and backward averaging matrix adopted is applied to eliminate the noise effect and estimate the accurate incident angle. As the same SNR with the one dimension method, the two dimension algorithm divide the difference matrix into several components to decrease the large estimation error bias induced by strong noise and incident source coherent effect. Simulation results show that the algorithm reduces the computation complexity and efficient estimation is achieved under the stationary and non-stationary noise condition as well.