The Reaserch Of Space Signal Processing Methods Appilied In Small Apterture Array

Along with the rapid development of science and technology, modern electronic systems with low cost, high mobility, high concealment and high adaptability are receiving more and more attention. Opposited to a large antenna array with tremendous physical size and huge economic cost,the small apterture array is much simpler,which not only can be built and managed with low cost but also can be set up and moved easily and flexibly. However, when some traditional techniques applied in small array, especially using the conventional beamforming, its performance is very poor, which is hardly to meet the requirements of direction finding and suppressing interferences in practical engineering. Signal processing methods with good performance are the key to the miniaturization of a large antenna array. Thus, how to make the performances of small array meet the actual project indicators by signal processing methods has broad applying prospect in practical engineering.Based on the actual radar and communication engineering background, in this topic, we mainly study superdirective beamforming and super resolution spatial spectrum estimation these two technologies, which can help the small antenna array to meet the actual project indicators. According to the characteristics of the small antenna array, many related problems aroud these two technologies need to solve. In this topic, we try to provide solutions to these problems and design small array, which meet the practical requirements.The content of the thesis is approximately as follows:1. As the small aperture array using superdirective beamforming exists many flaws, such as narrow bandwidth, low efficiency and extremely sensitive to array perturbation, we conclude the problems that need to be addressed. Based on the discussions on these problems, we provide a fundamental design precept for the miniaturization of HF receive arrays. In the disgn scheme, a novel superdirective beamforming method based on a constrained sensitivity factor is established as a core criterion. Provided that the given sensitivity factor is properly chosen, the proposed method not only presents a good tradeoff between directive gain, array efficiency and robustness against array error, but also provides an analytical solution, which corresponds to the constrained optimum directive Gain.2. Based on the adaptive array theory, we find that low side lobe level (SLL) and forming nulls can help to overcome the low array efficiency brought by superdirective beamforming. Therefore, for a small size array with accurately calibrated, its array efficiency corresponding to optimum superdirective beamforming method can be impoved to the demanded level by low SLL and forming nulls. Guided by this principle, we combined convex optimization techniques and the superdirective beamforming method based on a constrained sensitivity factor to provide a new design framework for superdirective receive arrays. In this new framework, SLL, nulls and sensitivity factor are not only performance indexs but also control variables. According to the different accuracy of array calibration, we can obtain a good tradeoff between directive gain, array efficiency, SLL, nulls and robustness against array uncertainty by choosing these three variable value, achieving the purpose of designing a optimal superdirective receiving array.3. Combined with the results of the previous study, we introduce the technique of spasrse representation to the problem of superdirective receive array pattern synthesis. Based on Euclidian1-norm constraint, Euclidian1-norm and Euclidian oo-norm mix constraint and Euclidian2-norm constraint these three different spasrse representation criterion, we popose three different superdirective beamforming methods. Among these three methods, the superdirective beamforming method based on Euclidian2-norm constraint could be viewed as a novel superdirective beamforming method based on constrained sensitivity factor, in which the SLL could be lower. It not only could be calculated simply, but also provides a considerable robustness against array error. Thus, it has less demand on the actual project condition.4. As the performances of high-resolution subspace-based algorithms are particularly sensitive to the prior information of the source number, the Signal-to-Noise Ratio (SNR) and the snapshot, we study direction-of-arrival (DOA) estimation methods without estimating the source number in this thesis. Accoring the different features of signal eigenvalues and noise eigenvalues, we reconstructs a new spatial spectrum, which is very similar to the SSMUSIC algorithm. In this new spatial spectrum, the prior information of the source number is unnecessary. Simliar to the above idea, we constructs a synthetic spatial spectrum with low computing complexity, which only need the inverse matrix of array covariance matrix and avoid source enumeration and eigendecomposition. From the computer simulation, it is indicated that compared with MUSIC method, the two proposed methods have better resolution performance, especially in small snapshot number and signal to noise ratio condition.5. As the performances of the second-order cumulants DOA estimation methods would be restricted by non gaussian white noise and the number of incident signals, to improve the performance of DOA in practice, we studied the four-order cumulants DOA estimation methods, which are suitable in arbitrary array. To reduce the heavy computation cost of four-order cumulants DOA estimation methods, we combined unitary transform and the idea of DOA estimation method without estimating the source number to provide a fast four-order cumulants method. From algorithm analysis and computer simulation, it can be found that the fast method not only can exihibt excelent performance in non gaussian white noise but also can estimate the incident signals while the number of sources are bigger than the number of array element.