Research On Task Assignment Of Multiple Mobile Devices In Power Internet Of Things Based On Edge Computing

With the emergence and development of the fifth-generation mobile communication network(5G),combining advanced information technology with the power system has become an important way for the power grid to realize the digital transformation of the power grid.However,the problems of energy loss and delay in 5G wireless networks limit its application range.In order to give full play to the advantages of 5G network and meet the communication performance requirements of largescale power growth in the field of power Io T,this paper constructs a ubiquitous cloud-side collaboration framework for power Io T based on the communication characteristics of 5G networks.,information reliability and other aspects,select the minimum weighted sum of energy consumption and delay,reduce data storage,communication and processing costs,improve the utilization rate of communication resources of 5G networks,and achieve customization and maximize utilization.Firstly,a greedy optimization algorithm based on Lyapunov is proposed for the resource allocation problem generated by the cloud-edge collaborative system for power inspection of mobile devices with energy harvesting capability.The optimization problem of dynamic minimization of the combined cost of mobile device delay and energy consumption is constructed under the gradual convergence of device battery power.Using Lyapunov’s dynamic optimization theory,the optimization problem is decomposed into three sub-problems of optimal local execution per time slot,offload execution and energy harvesting,and the optimal solutions of the sub-problems are obtained by linear programming.Then,by choosing the execution mode among local execution,offload execution and task discarding,the result of the minimum joint cost of the device’s delay and energy consumption is obtained.Finally,a greedy strategy program is designed using key-value pairs to adapt to multi-user multi-server systems.The simulation results confirm that the unloading rate can reach more than99.9%,and the service delay and system energy consumption can be effectively reduced under the condition of ensuring that the battery power of all devices is stable near the specified operating level.Then,aiming at the poor quality management of power Internet of Things on-site monitoring,an integrated edge computing system for risk operation supervision is designed.The edge computing technology is applied to the power Internet of Things risk monitoring field system,and a three-level collaborative architecture of cloud,edge and terminal is proposed.By taking terminal devices and edge servers as edge networks,a resource allocation model is established,and an edge network offloading algorithm based on distributed deep learning is proposed.The algorithm reduces the offloading decision optimization problem into a mixed integer programming problem and uses multiple parallel deep neural networks to generate offloading decisions.Simulation experiments show that the algorithm can effectively reduce the energy consumption and delay of the system,and can quickly make unloading decisions.