Research On The Distributed Relative Position Technology

With the ability of location awareness, the wirless devices can provide variable service for public, by acquiring the target position information. So the application field of location awareness became more and more widely. The key of location awareness was the distributed relative position technology. For achieving the sufficient research on the distributed relative position, there were two typical application field chosen as the research background in this paper, which were the cooperation of multi-airbrone and the through wall detection of the radar network.(1) There were two kinds of location problem under the situation of airborne cooperation, which were the position between two-airborne and multi-airborne.Under the research of the location between two-airbrone, the relative ranging measurement and the inertia navigation information(INS) were utilized to modify the relative error of inertia navigation systems for improving the accuracy of relative positioning. Then the relative ranging measurement was also employed to achieve the detection of the relative maneuvering state. Based on the results of maneuvering detection, the corresponding state model was chosen to provide the dynamic positioning between two airborne. Considering the property of the relative motion between airborne, this paper showed the modified different two way ranging protocol, which was used to estimate the model of the ranging. Then the cost function was built according to the relationship between the ranging error and the position of two airborne. At last, the cost function and the Lagrange multiplier method were combined to achieving the positioning between two airborne.Under the research of the location between multi-airborne, this paper presented multiple hypothesis least-squared algorithm to solve the under-redundant of relative measurement in the relative position between three aircrafts. Firstly, the multiple hypotheses were established based on the error model of INS. Then the different least-square estimate results of the relative position between aircrafts can be obtained under each hypothesis. Finally, the dynamic relative position can be acquired according to the Bayes's weighting and filtering. In additional, this paper introduced a kind of linear fusion algorithm to improve the accuracy and stability of estimation result by the optimal linear estimating model and the different estimate results from three aircrafts, the correlation of which was unknown.(2) There were two kinds of location problem under the situation of through wall detection of radar network, which were the self-location position of radar network and the multi-targets location and tracking by through-wall radar.Under the research of the radar network self-location, this paper presented the range weighted location algorithm to eliminate the shadowing effect of the wall to the ranging and location. Firstly, the cost function was estimated based on the geometric relationship of the ranging in radar network. Then the sequential quadratic program was employed to estimate the ranging error, which was used to modify the ranging information. At last, the self-location of radar network was finished by the modified ranging.Under the research of the multi-targets location and tracking by through-wall radar, the data association and the dynamic tracking of multiple extended targets. For the "measurement-measurement" data association problem, this paper presented the cluster algorithm based on the quality fuction of cluster to achieving the data cluster in signal radar. Then the two level m-best S-D association algorithm was presented. In this algorithm, the two level processing architecture and the m-best S-D association was combined to remove the effect of ranging error and realize the data global association. In additional, this paper presented the "measurement-track" association method based on tracking of region, which was combied with the sequential monte carlo to realize the joint detection and tracking of the number and the state of multiple extended targets.