Study On Methods Of DOA Estimation Based On Array Calibration In Complex Noise

Array signal processing is a new field of signal processing, and it is a very important content of modern signal processing. Array signal processing has wide engineering application such as radar, sonar, weather predicting, land and ocean exploring, seismic and biomedical signal processing etc.In the actual project application, because of various error inevitably, the actual array flow will often be a certain extent of deviation or disturbance,.In particular, the high-resolution spatial spectrum estimation algorithm performance would have severely deteriorated, and even failure.Therefore, the array error has been the high resolution spatial spectrum estimated that the technology moves toward practical a bottleneck.With the continuous development of science and technology', signal processing theory and methods to a higher level expansion has been the subject of dealing with the assumption that signal from the noise and Gaussian,a smooth transition to non-Gaussian and non-stationary signals from the linear system Nonlinear systems expand to non-causal , non-minimum phase, time-varying system, and the complex system of intelligent signal processing has become the focus of modern signal processing.Not only does it require its comprehendsive, rapid, flexibility, reliability,it alsodemands super-resolution processing,dealing with non-stationary and non- Gaussian treatment.The correlation function can only describe a random sequence and second digital features, although with a second order to describe the characteristics of digital signals in the Gaussian distribution function group has been adequate, but to use it to describe the non-Gaussian signal is not possible, Noise and often non-Gaussian noise; autocorrelation function contains only the amplitude information on the phase information is blind, and certain occasions, the phase information is more meaningful than the amplitude information from the point of view of stochastic processes, from high Andean signal is not a non-minimum-phase system identification, therefore, based on second-order statistics of the non-Gaussian method can not solve the non-minimum-phase signal modeling non-Gaussian and non-minimum phase systems, such as identification; correlation function in the characterization random sequence between themselves and random sequence in time when the relevance of, seems to have been insufficient. This requires people to consider new tools to describe the process for the distribution of the random group, and higher-order statistics in the high-end of this spectrum is arising in the context.At present the mass of method used in array signal processing domain is based on supposition of the white Gaussian noise and the steady signal. But in practical environment, the white Gaussian noise supposition may not always be true, and many signals exited are not steady. For example, many man-made signals ,such as BPSK, FSK, AM signals, exhibit in the cyclostationarity, and LFM signals used in radar applications. As a result, the performance of subspace-based DOA and polarization estimation techniques may degrade when dealing with non-stationary signals.This article focus on the array error correction and the higher order cumulant, the circulation statistics theory, for different background and the noise encountered in the actual application environment,the error correction and the complex noise inhibiting phase union, proposed that the different algorithm, enables this article the research to have the important theory significance and the practical application value.In this paper, the main work and innovation can be summarized as follows:1. First introduced the array signal processing's some elementary knowledge, and then introduced the array model error, analysis the cross coupling adjustment algorithm of array element in the presence of mutual coupling, and through computer simulation are given the existence of the array of conventional error signal source DOA estimation algorithm based on performance and the impact of second-order statistics of the mutual coupling correction algorithm performance analysis.2. Against Gaussian Noise background, bsded on the array of mutual coupling error model, using of higher-order cumulant algorithm, popularily appling the toeplize thinking of the mutual coupling correction algorithm, proposed a new method based on the accumulation of four bands each other coupling correction algorithm, which is the first innovation this point. The simulation of Chapter IV is fully explained: The algorithm is superior to inhibit the ability of Gaussian colored noise in the background noise Gaussian color to achieve better performance, and based on the accumulation of four bands mutual coupling correction algorithm Performance is superior to the four bands MUSIC algorithm.3. Against non-stationary noise background, bsded on the array of mutual coupling error model, introducing the cycle statistics theory, derived cycle MUSIC algorithm.based on the second cycle moments, promoting the use of second-order correction algorithm, proposed a new econd cycle of mutual coupling correction algorithm against non-stationary signals, which is the second innovation this point. The algorithm in the presence of mutual coupling can be very good separation of the Cyclostationary different frequency signal to the non-symmetrical distribution in the space of background noise to estimate accurately the source of DOA, and its estimate of the mean square error in the letter Noise ratio greater than 10 dB at constant zero, it is estimated that a very high accuracy.4. In circulation second-order statistics foundation, also unifies the higher order cumulant and the circulation statistics, proposed that one kind adjusts the new algorithm based on the fourth-order circulation cumulant's cross coupling, which is the third innovation, enabling the error correction algorithm adapted to arbitrary complex background noise. The method of both high-end and the cumulative statistics of the merits of recycling, and make up for a cycle of inter-related MUSIC algorithm can not colored background noise cyclostationarity the DOA estimated deficiency, the simulation results of Chapter VI of the proposed certification the effectiveness of the algorithm.