The Design And Realization Of Direction Finding Based On Specific Nonuniform Linear Array

Direction-of-arrival (DOA) estimation is an important researching branch of array signal processing. It has been widely used in many applications, such as radar, sonar, and communication systems. In order to avoid angle estimation ambiguity problem, the sensor spacing of the Uniform Linear Array is generally set no more than half the wavelength of the source signal. Thus, the array aperture is fixed if the sensor number is chosen. In the application of satellite receive antenna,H-F radar antenna and electric astronomy interference array, source signal number is large and target refresh fast. So, these systems always require high resolution and large array aperture. Comparing with the Uniform Linear Array, Nonuniform Linear Array having the same sensor number has larger array aperture, which results in the better performance in resolution. Thus, this paper investigates Nonuniform Linear Array based DOA estimation method to satisfy the requirements of spatial target detection.This thesis studies the Nonuniform Linear Array based DOA estimation problem conjugate with specific scientific research item. Firstly, the Seven-Cell Nonuniform Linear Array was analyzed, then the polar diagram,distinguish ability,multi-signal division of the array was studied. Secondly, the Nonuniform Linear Array based the direction finding design proposal was discussed and then confirmed. Some parameters such as the Snapshot, the step length of peak search and the disposal data type were fixed through several computer simulations.After researching the characteristics of the MUSIC algorithm, the DSP processing platform was set to realize the algorithm. In the design, the complex signal process in the MUSIC Algorithm was transformed into the real area so as to realize the algorithm with DSP processor. Four modules were designed separately. They are the module of data covariance matrix calculation,covariance matrix decomposition,signal number estimation and spectrum peaks searching. By descending the complexity of each module, the real-time performance of the algorithm was increased. At last, after summarizing the four optimizing modules mentioned above, a optimum way to realize the MUSIC algorithm can be got.Finally, the DOA estimation accuracy and the signal processing time were tested. And the simulation results of test show that this design proposal can satisfy the requirements of the project.