Research On Improved Range-free Localization Algorithm In Wireless Sensor Network

Wireless sensor network has the characteristics of low cost, low power consumption and self organization. It has been more comprehensive used in the fields of military, precision agriculture, medical security, smart home, environmental monitoring and other fields.In the field of wireless sensor network research, node localization is one of the important technology. The location of the sensor nodes are randomly distributed, only after the location information of nodes is determined, data were collected through the wireless sensor network has application value.The non ranging algorithm has the advantages of low cost and small energy consumption, which can be determined by the location information provided by the network. As a result, the application of Approximate algorithm in wireless sensor networks is widely used, which mainly includes DV-Hop algorithm, centroid localization algorithm and Approximate Point-In-Triangulation Test(APIT), etc.. In this paper, we analyze and improve the non ranging algorithm in wireless sensor networks.To address the problems of Approximate Point-In-Triangulation Test(APIT) algorithm in Wireless Sensor Network requiring high anchor node density, Out-To-In or In-To-Out error judgment, nodes failure and so on, an improved APIT localization algorithm based on Wireless Sensor Network was proposed. The method based on directional searched node and combined with RSSI ranging calculation by Gaussian fitting method was used to deal with error judgment. Besides, a geometrical method was used to process the failure nodes. The simulation results show that the improved localization algorithm performs greatly better, both in positioning accuracy and number of error judgment, than the traditional APIT localization algorithm and the modified localization algorithm of APIT based on mobile anchor node.To overcome the low positioning accuracy caused by the imprecise estimated distance and the accumulative location error, the DV-Hop localization algorithm based on wireless sensor network is optimized. This method, taking the summation of the beacon node average jump distance based on the weighted error correction as its average distance, was used to reduce the average jump distance error. Besides, an improved least square method was used to process the accumulative location error. The simulation results show that the improved localization algorithm performs greatly better, both in distance estimation and positioning accuracy, than the traditional DV-Hop localization algorithm, DV-Hop algorithm of weighted least squares and the improved DV-HOP positioning algorithm based on one-hop subdivision and average hopping distance modification.