Evolution Models Of Wireless Sensor Networks Based On Complex Network Theory

Wireless sensor network(WSN) is a multi- hop network formed from many self-organizing sensor nodes. It is one of the important support technologies in the field of Internet of things. WSN dynamic self-organizing evolutionary algorithms are the basis of many core technologies(such as coverage and connectivity, topology control, route optimization and other technologies). At present, the study of the topological dynamic characteristics of WSN by the complex network theory is increasingly attractive.The article firstly introduces some important concepts and important models in the field of complex networks theory, and discuss the same points of the topological characteristics of WSN and the complex network theory. In addition, two new methods are proposed to build WSN on the basis of the complex network theory and to achieve the purpose of localization in the new model. The main findings are as follows:(1)A new WSN evolution model with the small-world characteristics is proposed.The limited energy of nodes is inescapable "stumbling block" during WSN research. In view of this situation, a WSN evolution model with small- world characteristics is constructed by the introduction of "super nodes”. The super nodes will create some hyperlinks as the reliable shortcuts of the communication between the common nodes and the sink node. Among them, the super nodes have infinite energy, huge storage capacity, and strong capability of data processing. Thesis focuses on the optimize deployment scenarios of common nodes, super nodes and the sink node. Simulation results show that the addition of an appropriate amount of super nodes in a reasonable position to construct hyperlinks can effectively reduce the average path length of the network, reducethe energy cost of communications and increase the survival time of the node, so as to enhance the overall WSN life.(2) A WSN evolution model considering the importance of tasks is proposed.In the WSN, the new nodes are more likely to build linksto the nodes which have more surplus energy. And the power of the node is limited, sothe linksare established only between nodes within a radius of communication. The evolution model considering the importance of tasks that nodes are carrying for wireless sensor networks is proposed with taking the phenomenon into consideration. One of the most important differences between this model and other models is that the model takes the improvement of fragility of the network into consideration at the beginning of the design: balancing the energy consumption and the resilience against deliberate attacks by setting up the value of k. With mean- field theory, the derivation shows that the degree-distribution of this model is between the distribution of random networks and scale-free networks. The simulation study shows that: an appropriate increase of the scale of the local-world can greatly improve the efficiency of data transmission; Select a appropriate value of k depending on the specific situation can extend the life cycle of networks and have a good resilience to the malicious attacks.(3) Achieve the purpose of localizations in specific models.In a detailed analysis of the advantages and disadvantages of the existing classical localization algorithm of WSN, DV-Hop localization algorithm is chosen to achieve a localization of the target node in the two new proposed models. The location information of the super nodes and the sink node of the model based on the small-world model is known, so they can be directly set as the anchor nodes.In the model based on scale- free networks, anchor nodes are evenly set to achieve localization. The final simulations show that DV-Hop localization algorithm has a good performance and few errors in the two models.