Modeling Of Distributed Source And Research On The Estimating Algorithm Of DOA

The technic of array signal processing attracts more and more attention in the recent research. An important part of the research field is the estimation of DOA(direction of arrival),in which there is a hot research area named the estimation of DOA for distributed sources. The reason for the emphasis located on the 'distributed' is that the past research is always under an assumption that the sources are independent points even though considering mutiple-sources. But the actual form of the object is not a point especially when the reflector is very large or the observer and the object is close. It is not a real way that we keep on simulating the object with point model and there will be bad influence in the research of algorithm. The main task of the thesis is establishing distributed model based on the character of distributed sources and researching the method of DOA estimating. The main work and achievements are summarized as follows.1. Four distributed source models are introduced. There are several different modeling methods which considering different character of distributed source. They can explain each other although they are different. In other words, each of them can be expressed by other ones.2. Present estimating methods are sorted in three kinds, which are MUSIC-method, approch-method, ESPRIT-method. The three kinds of method are basic way for estimating the DOA of point sources. At first, the theory of each method is illustrated. Then the performance of different methods is compared after simulating many times. Finally, summarize the conclusion.3. The performance of different methods in different environment is explored. The difference between the methods based on the assumption of point model and distributed model is compared.4. The model of distributed sources with mutiple bright-points is established. Then present a way of establishing model for asymmetry distributed source.5. Improve some methods of estimating the DOA for distributed sources. Compare the performance of improved methods and unimproved methods. Apply the methods with actual data acquired in the experiment and show the result of...