Research On 2-D High Resolution Direction Finding Algorithms

High resolution direction finding through array signal processing has developed in recent decades as a spatial processing technology which is widely applied in electronic reconnaissance, communication, medicine and other fields. Based on the summary about the problems and the main algorithms of high resolution technology, the dissertation researches on the two-dimensional subspace DOA estimation algorithms and two-dimensional coherent signals DOA estimation techniques, and gives an effective two-dimensional high-resolution direction finding system solution. This main work of the dissertation is as follows:The different mathematical models under the different sources and the direction matrix of the commonly used arrays are given in an ideal physical environment assumption. The number of the sources is estimated by the information criterion to determine the signal subspace and the noise subspace in subspace algorithms.MUSIC algorithm and ESPRIT algorithm have wide range of applications in the two-dimensional DOA estimation, therefore theoretical analysis and simulation comparisons are made. The MUSIC algorithm is chosen as the main research focus because of its robustness and practical.The extended virtual array finding algorithm is proposed for the array physical restriction which extends the second-order MUSIC algorithm to the high-order cumulants, making the number of the resolved sources is greater than array elements, effectively suppressing Gaussian noise, and increasing the array resolution.MMUSIC algorithm based on the uniform linear array can solve the coherent sources estimated problem of the traditional method, so the research and simulation of this algorithm are made. The UCA is converted to a virtual ULA in the pattern space because of its non-Vandermonde structure, then the coherent sources estimation problem is solved with MMUSIC algorithm, while, it can only be one-dimensional azimuth search; To this problem, the simulation of smooth moving decorrelation algorithm which can be applied both in ULA and UCA is made and makes the outstanding 2-D DOA estimation performance. Based on the in depth analysis of the above algorithms, the dissertation gives a 2-D high-resolution direction finding system solution under the topic index. The solution takes the estimation of sources number, source coherence, the array structure, virtual array expansion and the performance of algorithms into account, and gets the best DOA estimation simulation result in the system.