Evolution Models Of Wireless Sensor Networks Based On Complex Network Theory

Wireless sensor network, as one of the important foundational technologies for Internet of things, is considered as multi-hop wireless self-organizational network system with a large number of tiny sensor nodes. It fuses multi-disciplinary technologies such as sensors, embedded information processing, wireless communications, distributed computing processing and so on, and can integrate the logical information with the physical world to achieve the real "pervasive computing" model. Topological generation mechanisms and self-organization evolving algorithms are the important basis of the key technologies of wireless sensor network such as coverage and connectivity, topology control, located routing. The research on topological dynamic characteristics of large-scale WSNs by complex network theory has become a hot topic.This thesis firstly describes several basic topology evolving models of complex networks and their typical statistical characteristic parameters, comprehensively analyzes the topology characteristics of WSNs and their features of complex network, then it surveys previous researches on WSNs topology by applying complex network theory. Following those researches, this thesis studies the generating mechanism and self-organization evolution models in real wireless sensor network to describe and optimize the topology of WSN. The main results of this dissertation are as follows.(1) A new topology evolution model by selecting the superior and eliminating the inferior mechanism based on energy-awareness for WSNs is proposed.Node energy is an important factor that affects on topology of WSN. Based on the dynamic evolution in WSN is closely related to the node energy, we introduce the preferred attached mechanism and the anti-preferred deleted mechanism according to node energy and present a topology evolution model by selecting the superior and eliminating the inferior mechanism based on energy-awareness for WSNs. By using mean-field theory, the degree distribution of this model involves into a scale-free state, and simulation experiments verify correctness of the theoretical analysis. Numerical simulations are conducted for investigating how different node energy distribution in the network affects topology characteristics and network performance. The simulation result indicates that the node with more energy carries more connections. When nodes energy is more heterogeneous, the network is better clustered and enjoys higher performance in terms of the network efficiency and the average path length for transmitting data. The model can be used to enhance the understanding of the evolving mechanism in sensor networks.(2) A new local-world evolving network model with energy-awareness is proposed.Sensor nodes can only communicate with their neighbor node because of their limited energy and power in WSN. This shows that the concept of local-world connectivity exists during the evolution of the wireless sensor network. From the perspective of local-world evolution, the thesis proposes a local-world evolving network model with energy-awareness for WSN. Theoretical analysis and simulations indicate the degree distribution of this new local-world evolving network model represents a transition between that of the random network and of the scale-free network. Numerical simulations are conducted for investigating how different local-world scale affects topology characteristics. Results show that increasing the local-world scale a little can improve the efficiency of data transmission. The robustness and fragility of the local-world network model also display a transition between the random and the scale-free ones. The model effectively approachs the true state of sensor networks and provide the usefule exploration on energy-saving application for the sensor networks.(3) An energy-effcient self-organization algorithm with heterogeneous connectivity is proposed for wireless sensor networks.Heterogeneity of node energy is a common phenomenon in wireless sensor networks. In such node energy heterogeneous sensor network, how to balance the energy consumption is the key problem on extending the lifetime of sensor network system. An energy-effcient self-organization algorithm with heterogeneous connectivity based on energy-awareness is proposed. Each sensor node in the network adjusts its own transmission radius based on the energy of the node itself and the residual energy information of its neighborhood nodes during the constructing and operating phase. Thus heterogeneous network topology, in which the nodes can choose different transmission radius, is formed. Compared with the homogeneous network, in which the node carries the same radius, simulation and analysis are conducted to explore the topology characteristics and robustness with different node energy distribution. We find that degree distribution shows the scale-free property in the heterogeneous model. The proposed network model enjoys higher efficiency for transmitting data, higher robustness under node random failures, and longer network lifetime than those in the homogeneous ones.(4) A wireless sensor network model with small-world concept is proposed.The small-world characteristic exists in many complex networks. For the energy efficiency issue in WSNs, a wireless sensor network model with small-world concept is proposed by the method of adding super nodes to form super links. From the perspective of a complex network theory, simulations analyze how super nodes affect the energy efficiency of the network model. The results show that the performance of network has evidently improved and the energy efficiency has also increased when sensor network has been appropriatly added a few super nodes.