Research On Sensor Management For Multi-Station Collaborative Localization

The passive sensor has good concealment and anti-jamming performance since it never emit electromagnetic signal.In passive sensor networks,the performance of target localization and tracking depend on not only sensor measurement errors,but also relative sensor-target geometry.This paper focuses on the passive source localization system and the multi-static passive localization system.The accuracy of target localization can be improved by sensor selection of passive sensor networks on the fixed platform or path planning of for passive sensor networks on the moving platform.The main contents can be described as follows:1.The sensor selection problem for multi-station passive source localization is developed.Firstly,the Cramer-Rao Lower Bound(CRLB)of target estimation error in three-dimensional space with bearing-only location is derived.Then an optimal sensor selection formulation with minimizing the maximum determinant of CRLB is proposed.The min-max problem can be converted into convex optimization problem by the semi-definite programming(SDP)method.The simulation results show that the SDP algorithm is effective and can improve the tracking accuracy comparing with the convex relaxation algorithm and random selection algorithm.2.The path planning for 3D bearing-only localization and tracking involving two mobile observers is presented.An optimal geometry analysis for 3D bearing-only localization and tracking involving two mobile observers is presented based on minimizing the trace of the CRLB.The optimal intersection angle and optimal elevation angle are developed under some specific assumptions in three dimensions,corresponding propositions are proved.Then the online optimization of observer trajectories to minimizing the trace of the CRLB subject to physical constraints is addressed.An interior point method is used to solve the nonlinear constrained problem.The effectiveness of the proposed algorithm was illustrated with several simulation examples.3.The observer trajectory optimization problem of the noncooperative bistatic passive radars is presented.Firstly,the Posterior CRLB(PCRLB)metrics of three typical localization systems including angel/bistatic range,angel/bistatic range rate and bistatic range/bistatic range rate are derived with known transmitter position.The observer trajectory optimization theory and method based on PCRLB is then developed.Secondly,the target state and the transmitter state are estimated with a sequential innovation modified EKF algorithm for the non-cooperative bistatic passive radar localization system with unknown transmitter position.Then the observer trajectory optimization with movable platform constraints is addressed.Finally,the simulation results illustrate the effectiveness of the proposed method.4.The joint path planning of transmitter and receiver for the airborme passive coherent location(APCL)is studied.Firstly,the PCRLB of angle/bistaric range localization in three-dimensional space is derived.Then the path planning model of joint transmitter and receiver based on PCRLB is formulated with the platform motion constrains and relative distance constrains between transmitter and receiver.A rolling horizon optimization method is developed to solve the problem.Compared with the path planning of receiver,the joint path planning of receiver and transmitter can further improve the performance of target positioning and tracking.