Spatial Spectrum Estimation And Its Application To Short Wave Direction Finding And Location

Spatial spectrum estimation is an important area in array signal processing, which can greatly improve the accuracy of angle estimation of spatial signal. Recently, with the development of digital signal processing technique, the theory and technology of spatial spectrum estimation has been widely applied, and has become the hot point of direction finding in telecommunication and passive location. In the field of shortwave direction finding, as shortwave signal has disadvantage of time-variant fading and interferences increasing, it became more complicated and difficult in direction finding and location. Under the circumstances of presentation above, spatial spectrum estimation provides a high precision method in direction finding and location.According to the characteristics of narrow signals including shortwave signal, DOA estimation is analyzed from the aspects of coherent signal, cyclostationary signal and two dimensional signal, then a suitable method for direction finding and location is given. The main contents can be summarized as follows:For the DOA estimation of coherent signals, a new method is proposed by exploiting maximum singular vector of an extended matrix. This method can decorrelate the coherent signals by constructing a Toeplitz matrix. Compared to the FBSS and ESPRIT-like method, the new method has better DOA estimation performance in low SNR.Based on the ESPRIT method, a Cyclic ESPRIT method is given for the angle estimation in bistatic radar’s MIMO array by exploiting the signal cyclostationarity, which constructs the cyclic autocovariance matrix of received signals, uses singular value decomposition, utilizes the consistency of the receiving and transmitting matrices, then achieves the automatic pairing of receiving and transmitting angles. Compared with the ESPRIT method, the proposed method can improve the DOD and DOA estimation performance in interferences.To estimate the 2D DOA of uncorrelated signals, with the L-shaped array, an improved ESPRIT method is proposed by constructing the extended matrix, which does not need spectrum peak searching and achieves automatic pairing of elevation and azimuth angles. Compared with the conventional ESPRIT method, the proposed method has better DOA estimation performance, and can resolve more sources. To reduce the computation complexity, an improved propagator method is also given. In addition, an improved root-MUSIC method is given, which exploits the extended matrix of signals received in X and Z axis, achieves automatic pairing of 2D angles, and has better angle estimation performance than the improved ESPRIT method and propagator method.To estimate the 2D DOA of coherent signals, with the cross-shaped array, a new 2D DOA estimation method based on the ESPRIT-like method is proposed, which exploits the symmetric property of two cross uniform linear arrays, and constructs two new Toeplitz matrices. Computer simulations are presented to show that the proposed 2D DOA estimation method for coherent signals has better estimation performance than the DOA Matrix method.A new shortwave spatial spectrum direction finding system is given. Based on SCA, the system adopts the idea of object oriented design, providing a kind of normal, open and interoperable platform. By loading different DOA estimation algorithms and softwares, it can dynamically deploy multiple functions of the system, realize the universalization and modularization of the hardware platform, and the transportability of the software.A new shortwave cooperative location method is also given in this paper. Exploiting the cyclostationarity of shortwave signals, the Cyclic DOA estimation algorithms are applied in the method, which can be effective to combat interference signals. Shortwave spatial spectrum direction finding system is connected with central controller, where the networking is flexible and the data transmission rate is very high. Adopting the weighed averaging algorithm, it can make use of uncorrelation between ionosphere channels, which reduces the sensitivity to the channel change, and improves the accuracy of location in complex environment. Simulation results show that the method has good performance and stability.