Study On Source Localization Method Based On TDOA And FDOA

With the continuous development of wireless positioning technology,more and more industries have used positioning technology,and the demand for accurate position estimation has become more and more urgent in various industries.Therefore,accurate positioning technology has attracted wide attention of scholars and positioning technologies based on wireless sensor networks have developed rapidly.When the unknown target to be located is stationary,only the exact position of the unknown target needs to be estimated,and when the unknown target is in motion,for example: navigation service,item tracking,real-time position tracking of person,etc.,in addition to estimating the unknown target we also need to estimate the speed.Commonly used positioning measurement models for stationary targets are mainly based on Time of Arrival(TOA),Time Difference of Arrival(TDOA),Received Signal Strength(RSS)and signal arrival angle(Angle of Arrival,AOA),etc.,these positioning models can only estimate the location of unknown targets.When the target is in motion,it also needs Frequency Difference of Arrival(FDOA)measurements to estimate the speed of unknown targets.The TDOA/FDOA joint positioning can effectively utilize the position and velocity information of the sensor,and obtain accurate position estimation.It also gets an estimate of the speed and improves the localization accuracy.For the problem of estimating the position and speed of an unknown target when the target is moving,the main research results of this paper are as follows:(1)This paper proposes an enhanced semidefinite relaxation method,the enhanced semidefinite relaxation method is an improved form of the existing semidefinite relaxation method,and the positioning result is more accurate.By using the existing intrinsic connection between the optimization variables in the existing semidefinite programming problem and transforming these relations into convex constraints,the tightness of the existing semidefinite relaxation method is effectively improved,so that the position and speed of the estimated unknown target are obtained.The accuracy reaches the Cramer-Rao lower bound(CRLB).The simulation results show that the performance of this method is superior to the existing methods in large noise.(2)This paper proposes a new iteration weighted least squares method based on TDOA/FDOA measurement.Through CRLB analysis,it is found that source location estimation using TDOA and FDOA is more accurate than using TDOA alone.Based on this fact,we use TDOA and FDOA to develop the nonlinear Weighted Least Squares(WLS)problem.The objective function of the WLS problem is the quadratic form,and the constraint is also quadratic.In this problem,the weight is first replaced by the unit matrix,then the solution is solved to obtain a set of solutions,and then the solution is brought into the weight matrix to update the weight and intermediate variables,and so on until the problem converges.The WLS problem is a Generalized Trust Region Subproblem(GTRS),which can find the global optimal solution.The method of this method is less complex and can save computation time.