Research On Node Location Algorithms For Wireless Sensor Network

Wireless Sensor Network(WSN)is composed of a large number of sensor nodes,which are characterized by low power,low energy consumption and small hardware structure.These nodes communicate in a multi-hop way and construct a ad-hoc network in order to perceive and monitor the objects in the coverage area in real time.WSN has been successfully applied in industry,military,environment and medical treatment in recent years.The successful development of many systems is based on explicit location information.As sensor nodes usually work in a complex and changeable environment and have a large number,the efficient and low-energy localization algorithm has become a hot spot of scholars.In view of the low positioning accuracy of the existing positioning algorithms,this paper mainly studies static node positioning,moving node positioning and anchor node dynamic path planning and positioning.The main research contents of this paper are organized as follows:1)A hybrid localization algorithm based on chaotic particle swarm optimization and Taylor algorithm is proposed.Firstly,TDOA positioning technology is introduced.Aiming at the disadvantages of Taylor algorithm such as being greatly affected by initial value and not easy to converge when solving TDOA equation,chaos particle swarm optimization is used to solve TDOA equations whose results are taken as initial values of Taylor algorithm and iterated to improve positioning accuracy.The mixed algorithm is simulated under the same conditions with Chan,Fang,PSO,chaotic particle swarm and Taylor method whose initial value is a real value.The results show that the difference between the hybrid algorithm and Taylor algorithm is not big,which solves the problem of Taylor's initial value difficult to select.2)An improved MCB localization algorithm based on RSSI for static anchor nodes is proposed.Aiming at the defects of large sampling area and poor sampling efficiency when using Monte Carlo method to locate mobile unknown nodes,some improvements are made.Firstly,RSSI ranging is introduced to narrow the one-hop two-hop anchor box and the sampling range.Secondly,in the filtering stage,an added filtering condition and the introduced method of virtual area are used to remove the samples that do not meet filtering conditions.In the resampling stage,the sample particles that meet filter conditions are taken as a new sampling area for sampling.Finally,in the position estimation stage,the gray scale prediction model is used to predict the approximate location of the nodes.If the predicted nodes are within the sampling range,the sample particles are weighted,and then Taylor method is used to further enhance the accuracy to obtain the final coordinates of the nodes.The simulation results show that the improved algorithm compared with MCB and RMCL algorithms under different parameters has higher accuracy.3)A genetic ant colony dynamic path algorithm based on special nodes and an improved weighted triangle centroid static node location algorithm are proposed.First of all,special nodes are selected by the number of neighbor node in the scope of nodes communication.Then the shortest path is obtained by genetic ant colony algorithm to locate other nodes.The positioning steps are as follows: firstly,the obtained RSSI value is processed by using gaussian filtering,and the collected anchor nodes are selected.Then,the weighted triangle centroid algorithm is used to calculate the coordinate value.Finally,the Taylor series expansion is carried out to obtain the final coordinates of the nodes.The simulation results show that compared with SCAN path and DOUBLE SCAN path,the path obtained by the algorithm proposed in this paper is shorter,and the improved weighted triangle centroid localization algorithm also improves the positioning accuracy and the rate of coverage.