Research On Cloud Robot Computing Offloading Strategy Based On Game Theory

Cloud robots have shifted their business to a new computing architecture processed by the cloud computing center,which has become a key research content in the robot system.Cloud robots have greatly improved work efficiency with the help of computing,storage,and network resources in cloud services,but relying solely on the central cloud is not very reliable.With the rapid development of the Internet of Things,the central cloud is facing a heavy network burden,and has been unable to meet users' requirements for low latency and high bandwidth.Therefore,an edge computing model has emerged.By lowering cloud resources and business platforms to the edge of the network,the user's physical location is closer to the business cloud.This reduces the delivery delay from the user to the edge server,and explores the inherent capabilities of the network to improve the user's service experience.Based on the edge computing model,this paper aims to improve the efficiency of edge computing offloading,reduce the energy consumption and delay of cloud robots,and design an improved game theory calculation offloading algorithm.Based on this environment,this paper studies the optimal strategy for cloud robot computing offloading.In order to maximize the efficiency of edge computing,an improved game theory calculation offloading algorithm is designed.The experimental results show that the algorithm can not only reduce the energy consumption of cloud robots,break the hardware conditions of cloud robots' computational complexity tasks,but also effectively reduce the time to complete the tasks,proving the feasibility of the algorithm.The main work of the paper is as follows:(1)The three-tier architecture of the cloud robot edge server is studied,so that the edge server can provide computing,storage,and network resources for the cloud robot;then the cloud service of the edge server based on OpenStack virtualization technology is constructed.Docker Driver for Nova is used to integrate Docker lightweight virtualization technology with OpenStack,and uses the ROS image container to provide a cloud service physical environment for cloud robot voice recognition tasks.(2)Aiming at the problem of edge server deployment,in the case where the position of the cloud robot is relatively fixed,the correlation between multiple cloud robots and multiple edge servers is studied,and a mathematical model of the access delay between the cloud robot and the edge server is established,and the mathematical model of edge server workload is established;Then,based on the improved K-Means algorithm,the optimal deployment scheme of edge server is studied.Finally,the purpose of reducing the access delay of cloud robots,and at the same time making the load between edge servers more balanced,and improving the quality of edge services.(3)Aiming at the problem of cloud robot computing offloading,in the scenario of multiple edge servers and multiple cloud robots,the energy consumption and time delay of cloud robots during the calculation of offloading are analyzed and a mathematical model is established.Then,an improved game theory calculation offloading algorithm is designed to transform the cloud robotic calculation offloading problem into a game form problem.By using a dynamic update offloading strategy,the Nash equilibrium state is achieved,and the optimal offloading strategy is finally obtained.Simulation results show that the method proposed in this paper not only reduces the energy consumption of cloud robots in performing tasks,but also shortens the average completion time of tasks and makes full use of edge server resources.