Study On Technology Of Spatial Spectrum Estimation Based On MUSIC Method

Nearly 30 years ago, as an important branch of signal processing, array signal processing had acquired a wide range of applications and rapidly development, it was widely used in communication, radar, sonar, seismic exploration and radio astronomy etc. The main elements of array signal processing includes beam forming technology, null forming technology and spatial spectrum estimation.Spatial spectrum estimation technology (direction of arrival estimation technology) is a new technology of airspace signal processing, the core of it is the super-resolution estimation to the direction of arrival (DOA) of the airspace signal, and the main objectives of it is to study the algorithms which can improve the azimuth estimation accuracy and azimuth resolution of the airspace signal in the processing bandwidth or which can improve the operation speed.In recent years, the various algorithms of spatial spectrum estimation had gained fruitful achievements, and the theory had improved steadily. Having read a lot of transactions in the area of interest, the author aims at the subject on spatial spectrum estimation technology and the various algorithms and makes efforts to study the problems below:1. The adaptive beam-forming algorithms are studied, the mathematic model of array antenna is analyzed, and the classification of the adaptive beam-forming algorithms are summarized, the time updating algorithms, which are divided of adaptive time equilibrium and spatial equilibrium, are also simulated and analyzed.2. The Delayed-add algorithm, Capon algorithm, Linear prediction algorithm, Subspace fitting method in spatial spectrum estimation technology are further studied and analyzed through the simulation. In addition, number of signal estimation theory, which is the foundation and key of spatial spectrum estimation, is also studied and simulated.3. MUSIC method and ESPRIT method, which are both belong to the eigens- tructures subspace algorithm of spatial spectrum estimation, are further studied and simulated. And according to the condition when the incident signals are coherent or highly correlated, two decorrelation method, spatial smoothing method and modified MUSIC method are studied.4. On the basis of the spatial smoothing method and modified MUSIC method studying, the author compares the estimation performance of the two methods in the condition when the incident signals are low SNR signals. The conclusion is: for the incident signals which angle between close and low SNR, the estimation performance of modified MUSIC method is better than the spatial smoothing method, and the modified MUSIC method do not need sacrificing the effective element number of array, it has little calculation amount, convenient to hardware realization.