Localization algorithm based on SVM-Data Fusion in wireless sensor networks

摘要

This Positioning process of the Wireless Sensor Network (WSN) nodes is interfered by multipath, multiple access, especially NLOS transmission effect, Basing on TOA/TDOA positioning technology, the article brings forward TOA/TDOA measurement data model, and improves on a location algorithm which bases on Least Square Support Vector Machine(LS-SVM). On one network with uniformly distributed nodes, node localization experiments show that SVM data fusion location algorithm can effectively reduce the effect of distance estimation error on positioning accuracy, minish coverage, loopholes and network cost.