Research On Resource Optimization Algorithms Based On Game Theory In Electric Internet Of Things

Due to the low efficiency of traditional power distribution network and the excessive use of fossil fuels which can cause global warming and other issues,it is very urgent to establish a new and efficient green power distribution network.Electric Internet of Things(EIoT)can integrate green power resources into energy distribution systems,and can improve its efficiency,reliability,and security through automation and modern communication technologies.Although the basic equipments in EIoT have advanced communication and computing capabilities,the resources of the equipments are limited.In the process of information flow,if the wireless resources and computing resources owned by the equipments cannot be reasonably allocated,the user’s quality of service(Qo S)will not be guaranteed.Therefore,it is necessary to study the resource optimization problem in EIoT,and game theory has been proved to be effective in solving the above problem.This thesis focuses on the resource optimization problem based on game theory in EIoT.A joint resource scheduling algorithm based on network formation game,a task offloading algorithm based on coalition game,and a cloud resource allocation algorithm based on hierarchical game are proposed.The main work is as follows.(1)A network formation game based joint resource scheduling algorithm(FJRSA)in EIoT is proposed.The algorithm forms a tree topology structure through the cooperation of adjacent smart meters(SMs)in EIoT.Through this structure,the data of SM is transmitted to the access point(AP)and the task on the SM is offloaded to achieve the purpose of minimizing the transmission and computing energy consumption,and making full use of the wireless resources and computing resources of SMs.Then the network formation game model is developed.The participants in the model are SMs.The strategy of SM is to choose the path to connect to AP.The cost function is SM’s energy consumption.The simulation results show that using the FJRSA algorithm proposed in this paper,the SM can quickly self-organize into a tree topology rooted at the AP;compared with the SM-AP algorithm where all SMs are directly connected to the AP,when the number of SMs increases,the FJRSA algorithm can significantly reduce the SM’s energy consumption and improve the SM’s wireless resource utilization.(2)A coalition game based task offloading algorithm(CGTOA)in EIoT is proposed.This algorithm investigates the problem of task offloading about multiple SMs in EIoT based on non-orthogonal multiple access(NOMA).The computing tasks of SM can be processed locally or offloaded to the AP server.By choosing an appropriate offload strategy,the purpose of minimizing SM’s energy consumption and delay is achieved.Then,a coalition game model is built.SMs are regarded as the participants.SMs that sharing the same sub-carrier form a coalition.A shared sub-carrier can be used for task offloading of multiple SMs.The simulation results show that the CGTOA algorithm can converge quickly.Compared with other existing algorithms,the CGTOA algorithm has a lower total cost function and a higher total utility.(3)A hierarchical game based cloud resource allocation algorithm(HCRAA)in EIoT is proposed.This algorithm investigates the problem of cloud resource allocation in the case of multiple service providers(SPs)with multiple users in order to maximize the benefits of SPs and users.A hierarchical game model is built.At the lower-level,evolutionary game is used to simulate the SP selection of residential users.At the upper-level,non-cooperative game is used to simulate the competition among service providers.Then,the upper and lower level can reach the Nash equilibrium(NE)and the evolutionary equilibrium(EE)are proved,respectively.Simulation results show that after multiple iterations,the hierarchical game model can converge to an equilibrium state.Compared with other existing methods,HCRAA has lower user costs,higher user revenue and SP revenue,and reaches a balance between supply and demand.