Direction Estimation Under Array Error Condition

Array signal processing is an important branch of signals processing. It has a wide-ranging application vision in such as radar, communication, astronomy, biomedical etc. Parameter estimation of spatial spectrum is an important research subject of array signals processing. The theoretical research has more and more attention in the academic community. Compared to the additive noise, multiplicative noise system is closer to the actual situation of the model. This thesis analyses the effects of multiplicative noise. The main contribution and conclusions of the dissertation are in the several aspects as follows:1,The theoretical of array signal processing and fourth-order cumulant are introduced. The extending character of fourth-order cumulant is analyzed MUSIC-like algorithm and virtual-ESPRIT algorithms based on fourth-order cumulant are studied. At last, the method of removing the multiplicative noise by forth-order cumulant is studied to give the direction of arrival estimation.2,Spatial spectrum estimation of the multiplicative noise model is discussed. The Higher-order moments of observation process is used to extract the statistical properties of the complex signal process of the multiplicative noise. Finally, independent and identically distributed complex zero mean multiplicative noise, independent and identically distributed complex zero-mean multiplicative noise, complex multiplicative colored Gaussian noise three cases are analyzed.3. The problem of two-dimensional DOA estimation is studied when the multiplicative noise and additive noise existing. Because of the cyclostationary signal widely used in engineering practice, Taking into account the advantage of the higher-order cyclic statistics which are inhibiting any smooth (Gaussian or non-Gaussian) noise and non-stationary Gaussian noise, and separating the Stationary from cyclostationary signal, and The digital characteristics of measurement noise is discussed by the first and second moments from the cycle, each cycle and third-order cyclic moment of the observation data calibrate the source array, finally reconstructing without noise covariance matrix space and estimating the DO A..