Font Size: a A A

The Research Of Node Localization Algorithm Of Wireless Sensor Network Based On DV-Hop

Posted on:2017-06-09Degree:MasterType:Thesis
Country:ChinaCandidate:X S ChenFull Text:PDF
GTID:2348330488962333Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Because of the wireless sensor network with low cost, low power consumption, strong concealment, real-time information and strong adaptive ability etc., it has been widely applied to precision agriculture, security monitoring, space exploration, intelligent transportation, health care and other fields. If cannot be determined the location of the data source, the data that obtained from network is meaningless for us. In addition, finishing the position location of the whole unknown nodes in Wireless Sensor Network, can not only provide namespace for network, but also provide technical support for the integrated management of the network, load balancing and automatic configuration of the network topology. Therefore, the design of the localization algorithm which applied to wireless sensor network node with low positioning error, high node localization rate, has been a hot pressing technology in the field of wireless research.DV-Hop algorithm is widely used in people's life, but this location algorithm also has some shortcomings in practical application. In this situation, in order to reduce the positioning error, and increase the node localization rate, the article combined with centroid localization algorithm, weighted DV-Hop localization algorithm and lest squares curve fitting technology to explore an improved DV-Hop node localization algorithm based on centroid localization. Firstly, the algorithm makes a rough positioning of the nodes by the centroid algorithm. Secondly, the weighted average hop of the unknown nodes can be obtained by weighted DV-Hop localization algorithm, which can be closer to the actual value. When calculating the distance between unknown node and anchor node, through the least squares curve fitting technology and the location of unknown nodes calculating by the centroid algorithm to choose anchor node as the reference anchor node, which is the closest straight line. Finally, the coordinates of unknown node are obtained by using the node calculation method. Through the Matlab7.0 simulation experiments, the average localization error of the improved DV-Hop algorithm based on centroid localization is 20% lower than the DV-Hop algorithm. At the same time, the average localization error of the new algorithm is 13% lower than weighted DV-Hop localization algorithm. Meanwhile, this paper proposes a method to setting the weight for the already positioned unknown node to improve the localization rate, when the number of anchor node in network is small. The weight is mainly composed of distance estimation error and reverse calculation error. In the end, combining this method with the improved DV-Hop algorithm based on centroid localization, which can improve the positioning accuracy, but also can enhance the node localization rate. Experiments prove that under the same proportion of anchor nodes, the node localization rate of the algorithm of unknown nodes to participate in the orientation is 10%~25% higher than the improved DV-Hop algorithm based on centroid localization. The main contents of this paper are summarized as follows:Introducing the background and significance of wireless sensor network positioning technology, and the current research status at home and abroad, summarizing the basic principle, performance evaluation standard, node calculation method of wireless sensor network node location, and then comparing and analyzing several range-free classic localization algorithm and classic range-based localization algorithm.After studying of DV- Hop algorithm in-depth, and studying the influence of positioning error caused by communication radius and the number of anchor nodes through the platform of matlab7.0. Secondly, the article analyzed the location error of each node in the process of localization, and pointed out that the localization algorithm is not stable.Due to the error of DV-Hop algorithm generated by calculating the average hop distance of unknown nodes and selecting the reference anchor nodes. In this paper, a new algorithm of unknown node localization is proposed by combining the centroid localization algorithm and the least square curve fitting technique. And then the paper verified the influence of the new algorithm's positioning results caused by anchor nodes ratio and communications radius with the help of matlab7.0.Another problem is that the number of anchor node is too small to locate all unknown nodes. In order to increase the number of reference nodes, reduce the node localization error and promote the node localization rate, this paper puts forward a method that unknown nodes can participate in next positioning. According to the weights that composed of anchor nodes' measurement error and reverse calculation error to judge whether to choose the unknown nodes to participate in the next position or not. Then, the method is combined with the improved DV-Hop algorithm based on centroid localization. Finally, the algorithm is simulated by Matlab7.0 simulation platform. The simulation results show that the simulation results show that the new node localization algorithm can not only greatly enhance the node localization rate but also can ensure the node localization accuracy when the number of anchor nodes is small or sparse.
Keywords/Search Tags:Wireless Sensor Network, DV-Hop, Least Squares Curve Fitting, Method of Unknown Node Participating in Orientation
PDF Full Text Request
Related items