Research On Localization Algorithm Based On Monte Carlo In Mobile Wireless Sensor Networks

Research about acquiring the information about network nodes position has become one of the critical technology in Wireless Sensor Networks(WSNs). With expanding the scope of sensor networks application, it is changing from static networks to dynamic networks about the sensor networks. Currently, localization technology about Mobile Sensor Networks(MSNs)has become the focus of concern. Mobility makes the networks acquire more position information, but it also makes accurate localization more difficult.Firstly, this thesis summarized the MSNs, including their development and application status, network structure, characteristics and the most popular research technologies, and then it put emphasis up on the meaning of localization technologies in MSNs. The next part of the thesis mainly detailed several classical range-free node localization algorithms, including centroid localization algorithm, convex programming algorithm and DV-Hop algorithm. In the end, it inferred Monte Carlo localization algorithm based on range-free, and this algorithm could adapt to mobility and low density of the nodes and made better positioning accuracy at the same time. So this thesis mainly focused on an improved localization algorithm based on MCL for MSNs.This thesis proposed Monte Carlo localization algorithm based on overlapping area for MSNs by discussing the idea about sample-adaptive Monte Carlo Localization Boxed algorithm(AMCB)and constructing node motion model. The algorithm was an improved algorithm based on the traditional Monte Carlo localization idea. The node motion model which was used to forecast the movement direction and speed of the mobile nodes, combined with AMCB algorithm was established. Finally, sampling area was the overlapping area which was formed from beacon box and fan-shaped area of the node motion model. In the end of this thesis, it analyzed the performance from the aspects of localization accuracy and sampling frequency compared with the former algorithms with MATLAB platform. Finally, the simulation results showed that the proposed algorithm was effective and feasible in MSNs, reduced energy consumption under the premise of maintaining the positioning accuracy and delivered better performance.