Multi-sensor Multi-target Positioning Method Research Based On Imaging Strategies

The Positioning System based on Multi-Sensor, which lays a large number of sensor receiver device on the ground and fuses the radar echo data of every receiver, obtains the information of the target on the air and achieve the goal of locating and tracking the target. However, because of the existence of multi-target, the Positioning System based on Multi-Sensor need to combine the echoes from multiple targets, to realize the separation of echo data from different targets, which is called the data association. When the number of targets and receivers is large, the computation is quite considerable. Aiming at this problem, we launched a new multi-sensor multi-target localization technology.With the support of a significant aviation science funds, the multi-target positioning problem can be referred to the common imaging strageties by considering the sensor network consisting of multiple sensors as a two-dimensional sparse array in this dissertation, can be solved by projecting the echo data into the three-dimensional geographic space. The important task and innovation features are listed as follows:1. Research on the basic theory of multi-target positioning based on sensor-network. A Sensor Network Model with one transmitter and multiple receivers is established for the multi-base positioning, the basic principle of the time difference positioning technology based this model is analyzed, and the advantage and disadvantage based on time difference positioning is discussed; The basic principle of radar imaging is briefly introduced; Aiming at the problem of large computation and not convergencing of equations caused by solving multi-target’s position, a traversal search strategy for the problem is studied and the principle and characteristic based on the projection-imaging strategy for the multi-target positioning is analyzed; The coherent-processing signal model based on the projection-imaging strategy is studied, and according to the problem of strong target echo submerging the weak target in the non-coherent case, an signal- accumulating model based on the statistical probability of the target’s echo is put forward.2. Research on the multi-target positioning method based on the geographic space projection-imaging. According to the principle of multi-target positioning technology based on the projection-imaging strategy, this paper proposes a geographical space projection algorithm, which avoids the process of solving the complex equations by imaging for the observation area to attain the multi-target information from image angle. Given the sparse nature of the layout for the sensor network and the space target, a greedy algorithm for extracting the target in the sensor network image is put forward, which realizes the purpose of eliminate the crosstalk of different goals and improves the positioning performance of the system by extracting the strong points in the image and eliminating the side-lobes according to the strong points.3. Propose a multi-target positioning method based on Bistatic Range Space Projection. Based on the limit of the system resolution, the irregularity and space-variant feature of the fuzzy area is analyzed; Aiming the problem of not reasonably partioning the monitoring area leading by the irregularity and space-variant feature of the fuzzy area, a Bistatic Range Space Projection algorithm is proposed, which maps the fuzzy area in the geographic space Projection into the regular fuzzy area in the bistatic range space projection by bistatic range mapping to avoid losing the targets in the geographic space projection. Combined with the greedy algorithm for reconstructing the target, through the simulation, this paper analyzes the multi-target positioning performance between the Geographic Space Projection algorithm and Bistatic Range Space Projection algorithm and the study is implemented on the factors of influencing the system’s stability and positioning accuracy.