Research On Routing Protocols Of Wireless Sensor Networks Based On Game Theory

We are now living in a modern society with rapid development of informationtechnology. It will be a world of IoT (Internet of Things) that intelligent applications arefilled in human’s daily life. As the crucial basic support layer, the research of thetechnology of WSN(Wireless Sensor Networks) is significant and essential. In this thesisthe structure of the WSN is firstly studied to achieve an intuitive understanding of WSN.Secondly, several major features of WSN are analyzed by a perspective of module, whichindicates that WSN is greatly different with traditional wire-linked network. Somerouting techniques such as Flooding Directed Diffusion, Rumor routing GBR IDSQ areintroduced in detail. Aiming at getting the inspiration and tracking the designer’s minds,deep research and analysis have been done and knowledge of improving a routingprotocol has been accumulated.After my research and analysis of the classical clustering routing protocol LEACH,formula involved in the paper and settings configured in the protocol are well studied.Several fatal problems in LEACH are found as followings:1.the way of node communication are found to be roundabout2.the selections of cluster head ignore the residual energy of the node.3.the node’s behaviors make no sense of the historical contribution.4.low efficiency of the rout choosing. It is those problems that hinder theperformance of the network of LEACH by reducing the node’s energy and TTL.A improved algorithm LEACH-SV based on Shapley Value is presented after theresearch of game theory and the study of the instances of game theoretic application onWSN. The newly-proposed protocol makes contributions in three ways:1.The selection of the cluster heads is changed. In the traditional way, the clusterheads are chosen randomly while our improved protocol takes the consideration ofresidual energy, node’s position and the historical contribution.Actually we select acandidate cluster head instead of a true head. As more factors are involved, LEACH-SVholds more advisability than the previous one.2.After candidate cluster head selection, nodes with high magnitude of those threeparameters are screened. A model of Bankruptcy Game is imported to solve the problems.For the sake of getting a solution, we calculate the Shapley Value and finally the networkconfirms the role as cluster head,sleeping mode and normal node. By doing all these, thecommunication of the nodes are tending to be more efficiency and the advantages of fairallocation by Shapley Value are fully revealed.3.In LEACH,nodes are all in the same setting parameters. A new communicationmechanism is designed by finding the minimum communication price of the distance of clusters and base station. A threshold of switch is designed to prevent the overload ofbase station.The simulations are implemented by the powerfully tool MATLAB. After strictdesigning and patient data collection, meaty results of the experiment are acquired. Bycomparing with LEACH protocol and DCHS protocol, in the aspect of node deadnumbers, LEACH-SV performs27%and19%better respectively. In the aspect of meanresidual energy, the point of inflection in LEACH-SV emerges140rounds and100rounds later than LEACH and DCHS respectively. The performance in sending packets isrelatively weak which LEACH-SV holds10times and1.8to2times to the LEACH andDCHS respectively. The distribution of the energy in LEACH-SV is more uniform sincethe mutations of the contour color are later. Delay of the network in LEACH-SV is10%slower (500rounds)than the LEACH and DCHS. The thesis proves that LEACH-SVperforms better in energy consumption and the node’s survival time than those ofLEACH and DCHS. However, LEACH-SV performs worse in some aspects. Theexperiment indicates that the initial goal of designing this protocol is achieved.