Research On Node Localization For Wireless Sensor Networks Based On Connectivity

As a new information acquisition and processing technology, Wireless Sensor Networks(WSN) have made such a great progress in theoretical research and practical application since the mid 90s of the 20th century when it began to rise. With the continuous progress of the practical process, the initial vision of WSN application is gradually becoming a reality and playing an important role in national defense, environmental monitoring, smart home, logistics and transport and other application fields.In Wireless Sensor Networks, network connectivity is the basis for reliable data transmission, and is also the necessary condition for localization technology, topology control, routing protocols and other technologies. Node localization technology is one of the key technologies of wireless sensor networks, and it is the basis for incident location reporting, target tracking, geographic routing, network management and other system functions. Therefore, the research on the node localization technology based on network connectivity in Wireless Sensor Networks is of important theoretical significance and practical value.The purpose of this dissertation is to discuss node localization technology applicable to Wireless Sensor Networks. Aiming at the Wireless Sensor Network connectivity and node positioning technology features, the research on the node localization technology based on network connectivity in Wireless Sensor Networks is carried out. The main contents include:(1) Study on connectivity in Wireless Sensor Networks based on boundary effects. From the concept of node effective communication area, the connectivity of border nodes and the whole connectivity of network are deduced by theoretical method. The approximate upper bound of k-vertex-connected probability under the influence of boundary effects in Wireless Sensor Networks are given. After that, through simulation research, the following aspects are further analyzed under the existence of boundary effects and eliminating the effects, including the probability of isolated nodes, the expectations of network average degree and minimum degree, the distribution of network connectivity, the expectation of network connectivity and the probability of k-connected, so as to explain the influence of boundary effects on network connectivity. The simulation value and theoretical value of k-vertex-connected probability under boundary effects are contrasted and the simulation result verifies the theoretical analysis. The study on connectivity in Wireless Sensor Networks based on boundary effects is of some significance and reference value for node amount and communication radius setting, and also laying a foundation for connectivity-based node localization technology.(2) A node localization algorithm (LAND) based on neighbor nodes distribution for distance estimation is proposed. First, a distance estimation algorithm (DEND) based on neighbor nodes distribution is put forward to estimate the best distance value of cluster member nodes to the corresponding cluster head nodes. Then, on the basis of algorithm DEND, the node localization algorithm LAND is proposed. There are four steps in LAND, including self localization of cluster heads, statistics on adjacent node information of cluster meber nodes, estimation the distance based on DEND and using maximum likelihood estimation method to localize unknown nodes. Finally, through simulation method, the effects of boundary localization area to the algorithm localization coverage, the influence of beacon nodes ratio and deployment form on the algorithm localization accuracy are analized, and the algorithm LAND campared with typical algorithm LAEP are discussed. The simulation results show that: LAND algorithm can achieve high localization coverage; LAND algorithm can achieve good localization accuracy even in the condition of low beacon ratio; The localization accuracy of LAND algorithm is higher when the beacon nodes are deploied in rectangular format than in triangle format; Under the same conditions, the localization accuracy of LAND algorithm is better than LAEP algorithm.(3) A virtual cluster based mobile beacon aided node localization algorithm (VCMBLA) is proposed. The foundation of VCMBLA are the multi-hop distance estimation method based on DEND and the mobile beacon path planning strategy based on minimal 3-hop 3-domiating set. The multi-hop distance estimation method based on DEND is for estimating the multi-hop distance between the virtual cluster member nodes and the cluster head. The main idea of the mobile beacon path planning strategy is to traverse the 3-hop 3-domiating set nodes through the Algorithm of Wandering Salesman Problem, so as to achieve optimized mobile beacon path and form a series of virtual cluster head nodes. The simulation results show that: the multi-hop distance estimation method based on DEND can maintain a smaller average distance estimation error when the network node density is high; VCMBLA algorithm can achieve a better accuracy when the network connectivity become better; Compared with the DV-Hop algorithm, under the condition of same beacon ratio, VCMBLA performs better localization accuracy when the network node density is higher; Based on the minimal 3-hop 3-domiating set path planning strategy, VCMBLA can achieve better localization efficiency than based on Gauss-Markov model.