| Localization is a key technology in wireless sensor network (WSN), while localization in indoor environment and three dimensional is very popular and will leading the future development of localization technology. This paper introduced some of the typical existing localization algorithms, analyzing their advantages and disadvantages. Then we proposed two improved algorithms named DCP3D DV-Hop (Degree of CoPlanarity based 3D DV-Hop) and LS3D DV-Hop (Layered structure based 3D DV-Hop), based on one of the typical range-free localization algorithm DV-Hop.For DCP3D DV-Hop, by analyzing the influence of different beacon node topology to localization results, a new concept "DCP (degree of coplanarity)" is proposed using some of the methods which can measure the shape of a tetrahedron. With measuring the DCP of every quaternion consist of beacons, the available quaternion can be elected for localization. In addition, the more hops between two nodes, the lower accuracy when ranging in DV-Hop, so we gave a multiple threshold constraint strategy to solve these problems. Simulation results show that DCP3D DV-Hop has a better performance in accuracy, bad node control and stability, but the proportion of nodes could be localized is lower, compared with DV-Hop.LS3D DV-Hop is proposed to solve the problem of three dimensional localization in buildings. Sensor nodes are deployed by layered structure in buildings, so it is hard to determine which floor the unknown node belongs when localization can’t provide enough accuracy. LS3D DV-Hop set floor to each node before deployment and use projection distance measurements to do trilateration, which reduce the influence of the network topology, besides, only 3 beacons is required in localization instead of 4. Simulation results show that improved algorithm performs better in accuracy, localization coverage, robustness and stability especially in network with few beacons and low density. |
PDF Full Text Request