Self-Localization Algorithm In Wireless Sensor Networks

Wireless sensor networks (WSN) is composed of a large number of sensor nodeswho have the capabilities of perception, computing and wireless communication. InWSN, monitoring the specific location of the incident is the premise of the sensornetwork for target identification, monitoring, tracking, and many other applications.Therefore, to study the low-power and low-error localization algorithm is an importantresearch direction in WSN.Firstly, the structural characteristics of WSN and sensor nodes are introduced inthis thesis. The location technology and the criterion of performance evaluation ofWSN are researched. Besides, the existing taxonomy for localization algorithms ismainly summarized.Secondly, a detailed analysis of the tradition DV-Hop localization algorithm andits positioning principle and positioning process are introduced in this thesis.Considering the localization error and the large amount of calculation of DV-Hopalgorithm, this essay got an improved algorithm called LNDV-Hop (DV-Hop in LocalNetwork). LNDV-Hop algorithm used the thought of local network to position theunknown nodes. LNDV-Hop algorithm has been improved in three aspects: Firstly, theaverage one-hop distance among anchor nodes was refined by means of least squaresmethod (LSM). Secondly, the unknown node’s average one-hop distance whichestimated by several anchors was considered by this algorithm and assigned differentweightings according to the hops to the unknown node. In the stage of coordinates’calculation, the unknown node used the local network which comprised by theunknown node and the anchor nodes near it to instead of the original global networkand used the maximum likelihood estimation to determine the coordinate itself.Simulating the improved algorithm with MATLAB, the results show that LNDV-Hopalgorithm reduces the localization error and at the same time reduces the amount ofcalculation.Lastly, the thesis researches the centroid algorithm and introduces its positioning principle and positioning process in detail. In order to improve the question that thedensity of anchor nodes has great impact on the localization error and coverage ofcentroid algorithm, this essay got an improved algorithm called an improved centroidalgorithm in low density of anchor nodes. Firstly, this improved algorithmdistinguishes the priority of unknown nodes according to the distance between anchorsnodes and unknown nodes. Then it uses the improved DV-Hop algorithm to locate theunknown nodes which have high priority and upgrades these unknown nodes to theanchor nodes. Finally, the improved algorithm locals the other unknown nodes whohave lower priority to use the method of centroid algorithm. The simulation resultsshow that this improved algorithm can reduce the localization error and improve thecoverage effectively under the condition of lower density of anchor nodes.