Task Offloading Strategy Of Cloud-edge Collaborative Cloud Robot Based On Game Theory

As data analysis,deep learning,and many other advanced technology developments,the researchers try to use robots to perform some more complex such as object recognition and real-time path planning work.However,due to some limitations in the field,robots are often limited by size and battery capacity,and they are usually not rechargeable in harsh environments.As a result,they can't host powerful hardware and often lack computing and storage resources.The emergence of cloud computing provides a solution to the problems of cloud robots.Cloud robot computing tasks can offload data processing tasks from resource-constrained robots to cloud servers through wireless access,which can effectively extend the service life of robots,solve resource-constrained problems,improve task processing speed and reduce task costs.In the process of cloud robot task offloading,there will be major problems such as how to rationally utilize the resources of central cloud and edge cloud,how to reduce the equipment energy consumption of the system,and how to shorten the average task completion time.In this paper,game theory is used to solve the problem of competitive game between robots,in order to maximize their own interests in the process of task offloading and find out the optimal offloading strategy.The research content of this paper mainly includes:1.The emergence and development of cloud computing and edge computing as well as the advantages and prospects of cloud edge collaborative services are studied.Centralized Cloud Computing(CCC)can provide enough resources in the process of Cloud robot task offloading,but it is difficult to meet the needs of robots in completing low-delay tasks.Mobile Edge Computing,(MEC)can make up for the short board of high delay in completing computing tasks,but its computing capacity and storage resources are limited.Therefore,the establishment of cloud-edge collaborative services during the offloading process of cloud robots can help cloud robots to complete computing tasks to the maximum extent.In this paper,different cloud edge collaborative task offloading models are established according to the different requirements of cloud robots.2.In view of whether the cloud robot computing tasks can be separated and offloaded,the cloud robot computing tasks are divided into two types: binary offloading that must be completely offloaded and partial offloading that can be offloaded,and corresponding offloading models of cloud edge collaborative system are established according to different task types.An integral,if a task is to build more edge cloud server and centralized the cloud and cloud server cooperate together cloud robot task offloading model,and use the game theory task offload algorithm,the cloud offloading the game into a robot task,dynamic update strategy offloading,find the Nash Equilibrium(NE)of the system,and finally find the optimal strategy.The simulation results show that for the indivisible computing tasks of cloud robots,the cloud-edge collaborative system can reduce the energy consumption of system execution tasks,shorten the average completion time of computing tasks,and make full use of system resources.3.Aiming at the computing tasks of cloud robots that can be partially offloaded,a cloud-edge cooperative task offloading model of single edge cloud server and cloud server cooperation is established.Will cloud computing task completion time and produce energy consumption as indicators of cost,set the price of different weights according to their own demand,multiple cloud computing tasks of the robot offload problem transformed into a model of multiple players involved in the game,we design a part of the task uninstall algorithm based on game theory(GT-PTO),Through the Nash equilibrium state under the algorithm,the optimal offloading threshold of the participants is found,so as to achieve the optimization of the total cost of the system.Simulation results show that this algorithm can reduce the energy consumption of cloud robot computing tasks,shorten the average task completion time,and greatly improve the quality of cloud side collaboration service for partially offloaded computing tasks.