Shi Wentao

Orthogonal waveforms design and performance analysis for MIMO sonar

In this paper, a Multiple-Input Multiple-Output (MIMO) sonar system which can deal with multiple targets is discussed, where orthogonal waveforms are transmitted simultaneously. According to the requirement of waveform orthogonality, the orthogonal frequency division pseudo-random signal (OFD-PR), orthogonal frequency division determinant signal (OFD-LFM) and the spread-spectrum code-based signal (Gold-BPSK) are designed. The Maximum Likelihood (ML) DOA estimator and CRB for MIMO sonar is described. The performance of MIMO sonar based on different waveforms is analyzed via experiments in an anechoic water tank. Simulation and experimental results show that the waveforms which we design possess the higher resolution for multiple source DOA estimation than its phased-array and they could resolve the two targets with a separation of 1/2 beam width at low SNR. The analysis of DOA estimates and experimental results show that the OFD-PR signal has much better performance than others.

Optimal sensor-target geometries analysis for Angle-of-arrival localization

The target position accuracy suffers from the bad relative sensor-target geometries for Angle-of-arrival localization techniques. To reduce the impact, the optimal sensor-target geometries in the 2D plane are investigated based on maximizing the determinant of the Fisher information matrix. To solve the optimal problem, the conditions that the local optimal solutions should satisfy are studied by discussing the first derivative and Hessian matrix. For two sensors with arbitrary distance to the target, it is shown that the optimal geometry is obtained if the two bearing lines are mutually perpendicular. The optimal geometry is studied for more than two sensors with the equal distance. Simulation results illustrate the effectiveness of the proposed the optimal relative sensor-target geometries.