Research Of DOA Estimation Based On Cyclostationarity

The estimation of direction-of-arrival is an important area in array signal processing. The methods canestimate direction-of-arrival of several sources in different directions simultaneously. Multiple signalclassification(MUSIC) algorithm which is the classical super-resolution algorithm is always used for theestimation of direction-of-arrival. With the development of researches people found that a lot of signalspossess the character of cyclostationarity. By exploiting the cyclostationarity of the signals, betterdirection-of-arrival estimation performances in signal selective direction finding, interference or noisesuppression and multiple signal processing can be achieved. Thus, the cyclostationarity-based DOAestimation techniques have been becoming one of the hot research topics in array signal direction findingareas.This paper is divided into three main tasks. After the introduction of the classical Cyclic MUSICalgorithm, Cyclic MUSIC algorithm and MUSIC algorithm are simulated together. Cyclic MUSICalgorithm has some advantages that MUSIC algorithm doesn’t have through computer simulation results.Two new methods based on the character of cyclostationarity are proposed for one-dimensionalindependent coherent signal sources. The first algorithm is based on the character of cyclostationarity andthe spatial smoothing algorithm. After the sub arrays are divided the mean of sub arrays’ Cyclicautocorrelation matrix must be found out to decorrelate. The other algorithm is based on the character ofcyclostationarity and the matrix modified algorithm which uses uniform circular array. The algorithmchanges uniform circular array to virtual linear circular array through the mode space transformationalgorithm. The mean of new output data’s Cyclic autocorrelation matrix must be found out to decorrelate.Computer simulation results show that two new methods are effective. But the first method has a smallamount of calculation, and the two method has the better performance in the case of low SNR. Two parallellinear array based spatial smoothing technique is also extended to the2-D case. A simplified-two parallellinear array are proposed for the problem of array element redundancy. Two method which uses less arrayelement can get the information of azimuth and elevation angle. Its validity is confirmed by computersimulation.