Two-Dimensional DOA Estimation Algorithm And Experimental Verification

Two-dimensional Direction of Arrival estimation is the important field on ArraySignal Processing and Passive directional technology. This paper base on the cloud-cloudlightning detection and warning system, do thoroughly research on the collection of2-dDOA estimation algorithm frequently used on direction finding systems. Taking thecomplex electromagnetic environment around the system into account, the paper showsthe key steps on how MUSIC works on our system. And a series of comprehensiveexperiment will verify the all aspect of performance of the system, such as real-time,stability, anti-jamming capability and so on.At first, this paper introduces the fundamental on MUSIC and Phase Interferometer.Then, we do a comprehensive research on the error model of MUSIC in three aspects: Thealgorithm theoretical error model, the error caused by frequency offset and array modelmismatch. A quick MUSIC algorithm which can be applied in a different array manifoldare proposed to reduce the algorithm complexity. On another hand, a phase interferometeralgorithm base on the eight element uniform circular array were used in our system andfinally proved its limitations in response to phase ambiguity. A phase difference matchingalgorithm based on spatial spectrum searching is proposed to solve the phase ambiguityproblem which all phase interferometer face.Based on the eight element uniform circular array, the paper introduced the key steps ofMUSIC algorithm designing in systems. At the first, from the study of the array manifold,we summed up the array manifold optimize storage method and the real number MUSICalgorithm based on the unitary transformation. Since the array manifold iscentre-symmetric, we introduce an active calibration scheme used in our system.Moreover, a stable and accurate source number estimation strategy are proposed based onexisting source estimation algorithm combine with the study of the actual data collectedin complex electromagnetic environment.At the end of the dissertation, several experiments result were presented which show thesystem performance on computing speed, direction accuracy, the stability and accuracy ofsource number estimation and the ability respond to multi-source signal incident. The Small-scale experiment with mobile sources shows the system omni. In a wide range offield, a4-station joined experiment shows the performance of the detection network’s jointfinding. The analysis on the error of the result will contribute to designing the positioningalgorithm.