Study On Three Dimensional Localization Technology Based On APIT In Wireless Sensor Networks

In recently years, wireless sensor networks is developing very fast with a verybroad prospects in many fields, such as military, medical, environment, industrial andso on. Node positioning technology is viewed as one of the supporting technologies ofWireless Sensor Networks’ applications, which is premise of target tracking, geographicrouting, network management and other system functions. The positioning accuracy isthe main measure of localization performance, directly reflecting the merits ofpositioning technology and the range of application. With no information of positioningscale collection, processing and transmission will become no practical Significance forwireless sensor network. Accurate location information, the lower energy consumptionand the distribute calculation is the base for a effective positioning mechanism and theyare the necessary factor for wireless sensor network. Currently, the research is very widefor two-dimensional location, but the case of application in actual three-dimensionalsituation is complicated. So it is very wide background of practical application andtheoretical basis of significance that do research to find a good accuracy positioningmethod for three-dimensional localization.With the deep research and development in wireless sensor network, APITalgorithm is viewed as one of the classic positioning algorithms, received the wideconcern from many researchers. In this paper, we focus on the study on thethree-dimensional range-free localization and do many works about the APIT-3Dalgorithm.In the process of APIT-3D，if we directly use the overlapping area of the testingpyramid which is compose by the unknown node and it’s neighbor beacons to locate, itis a bad positioning module with much error. This paper proposes a improving APIT-3Dalgorithm based on Fermat-Point fragmenting. We consider the testing pyramid isensured a bad range to locate and it can get a coarse positioning accuracy and this papertakes the help of the geometric reference point in algorithm. The positioning module ofAPIT-3D will be divided into pieces, which will be used to locate for the unknown nodeto improve the positioning accuracy. Then, the simulation results confirm that theproposed algorithm has great advantage in aspects of localization accuracy and networkcoverage in comparison with original APIT-3D scheme. Especially, the density of nodeis few and scattered, the effect is enough better. In the process of FM-APIT-3D algorithm, we have to use the PIT (perfectpoint-in-triangulation test) scheme and the bounding effect will come out, which causethe "In-to-out error" and "Out-to-in error" with a bad located error. Aiming the case, thispaper proposes a Volume-Testing scanning algorithm named VT, to improve theFM-APIT-3D. By adding the comparing advantage in PIT scheme, the unknown nodescan be accurately determined is in the testing pyramid or not. This method can removethe probability of the bounding effect in FM-APIT-3D. What’s more, this paper has alsooptimized the grad scanning algorithm in the aspect of computing rate. Finally, theexperiment data confirm that the improved algorithm has great advantage in aspects oflocalization accuracy in comparison with original APIT-3D scheme and relationalscheme respectively and also cut down the positioning time very notably. Such a fastaccurate three-dimensional localization method it is.