Research On Direct Position Determination For Emitter Signals By A Single Moving Observer

In modern warfare,the passive location technology is an important measure to acquire the deployment of the electronic system and access the battlefield situation.Conventional two-step localization methods usually estimate the intermediate parameters e.g.directions of arrivals,time difference of arrival and frequency difference of arrival in the first step,then determine the emitter positions using the previously estimated parameters in the second step.Compared with the two-step localization methods,Direct Position Determination(DPD)can obtain the emitter positions directly without estimating the intermediate parameters,which shows advantages in ability to locate the identical frequency signals with the same arrival time,higher localization accuracy at low signal to noise rate and no necessary to associate data.This process is achieved by modeling the relationship between the emitter positions and the intercepted signals,formulating the cost function only related to the emitter positions and solving the cost function.Since DPD with multiple observers suffers from the communication burden between different observers and the difficulty in time and frequency synchronization,this dissertation focuses on DPD with single moving observer.This method can estimate the emitter positions by processing the raw signals accumulated during the motion of the single observer,which contains the advantages of DPD and meanwhile avoids the communication burden and time and frequency synchronization.The major content of this dissertation can be concluded as follows:In chapter 2,the basic methods of DPD with Single Moving Observer(SMO-DPD)are studied.First,the narrow-band signals and coherent signals are formulated,and the intercepted signals are modeled based on different information parameters.Based on this,the SMO-DPD cost functions are formulated by different optimization methods.Then,the Cram(?)r-Rao lower bound(CRLB)for SMO-DPD is derived.Finally,computer simulations verify the correctness of the theoretical derivation and analysis above.In chapter 3,the high resolution and accuracy SMO-DPD methods are studied.First,SMO-DPD based on eigen space is proposed,and its superior resolution compared with the existed methods is verified by computer simulations.Then,SMO-DPD using a rotating array is presented whose CRLB is also derived.The computer simulations verified the localization resolution and accuracy are improved by the array rotation.Finally,coherent SMO-DPD is proposed for synchronous coherent signals,which significantly improves the resolution and increases the maximum emitter number that can be localized.In chapter 4,the SMO-DPD in the presence of sensor gain and phase errors is studied.The intercepted signal model and the cost function are first formulated.Then,the theoretical localization bias due to these errors is derived.Subsequently,the selfcalibrating SMO-DPD is proposed whose CRLB is also derived.The computer simulations verify its effect on calibrating the sensor gain and phase errors,but also expose its poor performance at low signal to noise rate.To solve this problem,SMO-DPD using a calibration emitter with known location and the corresponding computer simulations verify this method is effective.In chapter 5,the fast algorithms of SMO-DPD are studied.The computational complexity of the above SMO-DPD methods is first analyzed.Then,the fast algorithms based on multistage grid and Alternating Projection(AP)technology,based on Particle Swarm Optimization(PSO)with niche and based on PSO with niche and gradient class methods are proposed in sequence aiming to reduce the computational complexity of SMO-DPD using AP technology only and SMO-DPD using spectrum methods.The computer simulations verify the effect of these algorithms on reducing the computational complexity of SMO-DPD.