Self-localization Algorithms For Wireless Sensor Networks

In recent years, wireless sensor networks (WSNs) gradually into people’s horizons, itgets a lot attention and becomes a hot field. The positioning problem of the nodes is one ofthe key technologies of wireless sensor networks, we should not only understand theinformation of the area under test but also understand the information of location, someasurement information have some practical significance. For the present research situation,we are going to study the localization algorithm of wireless sensor networks.We researched the traditional range-free localization algorithm of DV-Hop, sinceDV-Hop algorithm has simple calculation process, easy to implement, adaptable and widelyapplication range. It also has some disadvantages, its positioning accuracy is limited and inthe anisotropic network performance is poor, aim at the two disadvantages, we suggestimprovements. First we used the Angle method to select anchors around the unknown node,taking advantage of the special properties of the triangle, selecting the anchor node routingholes smaller between anchor nodes to participate in the calculation, to improve the precisiondegree of the calculated average hop distance, and then improve the positioning performanceof the system by this way. DV-Hop is applied in2d plane algorithm, the method we putforward is applied in3d space, we refer to the parameter setting of coplanar and the hopthreshold to constraint the process of calculating, reduce the computational load of thenetwork, and exclude some larger brings to the positioning error of the anchor nodes, wecontrast the improved performance of the algorithm in the three-dimensional space usesimulation experiments, we find that the overall accuracy has improved.We divided the distributed localization algorithm into three steps through summarize theDV-Hop algorithm:1. Estimating the distance of the unknown node to anchor node.2.Though using information of estimate distance and the coordinates of the anchor nodes toestimate the location of unknown node.3. Using the mathematical method to operate theresults of the last step. Analysis the three steps above, we put forward a kind of combinationalgorithm. In the first stage of the localization, we still adopt the traditional method, in thesecond stage we use both Min-max and LI. Since the Min-max calculation process is simple,we used the Min-max to limit range, then we used LI algorithm to find the coordinates ofpreliminary results. In the last step, we adopt a method of iterative refinement to refinementthe calculation results, the advantage of SD algorithm is which can improve the positioningprecision of the results in case of guaranteeing the proportional constant of node. Bycomparing the location result of the traditional algorithm, positioning accuracy, powerconsumption, and performance ratio of the locating node of the combination of algorithmusually has better performance, we can say it is a kind of algorithm that has less dependencefor distance estimation accuracy, low dependence for the environment, the positioningaccuracy is higher and more stable in the two-dimensional plane locating method.