Research On Node Localization Technology In Wireless Sensor Networks

Wireless Sensor Networks(WSN), for its outstanding characteristics, such as low cost, fast net-setting, dynamic topology and multi-hops communication was widely used in the military, environmental monitoring, medical treatment,space exploration, intelligent household, commercial application and other fields. A number of sensor nodes were laid in the scene to observe physical phenomena, while in most cases, the data have significance on the condition of combining on location information, so the node localization technology as the core technology, plays an indispensable role in WSN systems. However, most of WSN positioning technologies have some limitations, such as depending on hardware peripherals, low precision, error accumulation and large energy consumption. Therefore, WSN localization technology combining with the introduction of different algorithm thoughts and theories was researched in this paper in the three directions including the static node localization, the mobile node localization and the mobile anchor node path optimization.The main research work of this paper is as follows:First of all, the paper has a brief overview on the characteristics of wireless sensor network and the methods of sensor node positioning technology,elaborate the background and meaning of this topic. The present researching situation of the three aspects: the static node localization, the mobile node localization and path-planning of mobile anchor node was displayed. And further introduce the typical localization algorithms and the methods of path planning. For the purposes of improving the positioning accuracy and reducing energy consumption, some improved algorithms were proposed.Second, the paper has a study on static node localization algorithm and a static node location prediction method based on support vector machine(SVM)was proposed. In the existing classical node localization algorithms, the range-based localization algorithms have high positioning accuracy, while they need to increase the peripherals and were highly affected by the outdoor environment; the range-free node positioning algorithm have high requirements to node distribution and can’t be applied to the occasion of node distribution. Furthermore, trilateration, triangulation and maximum likelihood estimation were used to calculate the position of the unknown nodes in most of localization algorithms, not only the position error is high, also have the problem of error accumulation. So a node location prediction method based on SVM was proposed, using the hop and distance information between the nodes to get hop-distance transformation matrix and combining with the principle of SVM prediction to forecast the unknown node position. This algorithm can get a large number of accurate positions of the unknown nodes in the cases of asmall number of anchor nodes, have the advantage of low cost and high positioning accuracy.Third, an algorithm for mobile node localization based on FOA-MCB was put forward in the situation that the unknown node is mobile and anchor node is static. As the widespread application demand of WSN, a new research direction on localization technology for mobile nodes was created, and the traditional localization algorithm for static node is not suitable for dynamic networks. MCL algorithm creates a new direction for the research on mobile node localization algorithm, and some improved algorithm based on MCL proposed, such as Dual-MCL, MMCL and MCB, though the positioning precision is improved for these localization algorithms but limited. Therefore,a kind of mobile node localization algorithm based on FOA-MCB was put forward in this paper. Considering the speed and direction of node in last moment has an influence on moving speed and direction of node in the current moment, the continuous correlation model be used to plan the movement trajectory of unknown nodes; The MCB algorithm is used to estimate the position of unknown node, through the estimated distance and the measure distance between anchor node and unknown node to get the fitness function of fruit flies optimize algorithm(FOA), to search the optimal location in the solution space. The algorithm decreases the positioning error in the form of iterative optimization and improves the positioning accuracy.Finally, a new path planning method for mobile anchor nodes based on breadth first search algorithm(BFS) and modified fruit flies optimize algorithm(MFOA) was presented. Mobile anchor node path planning is one of the important auxiliary means of mobile anchor node localization. In the static method of path planning, the anchor nodes usually scan the preset trajectory,as a result of the redundancy mobile of anchor node and existing failure nodes when the nodes have uneven distribution; In the mobile path planning methods,the anchor node mobile random, when the pending area is complex, as a result of redundant on radio location and poor positioning performance. So the problem of path planning was turned into graph traversal, BFS was used to pick up the virtual anchor nodes as the localization of mobile anchor nodes,further the MFOA was used to optimize the path of virtual anchor nodes and get the optimal moving path of anchor nodes. The moving path of anchor nodes can change adaptively according to the distribution of the unknown node. Getting the optimal moving path of anchor nodes to make a better matting for the next study on mobile anchor node localization technology.