Task Offloading Strategy In Edge-Cloud Collaborative Environment

The quantity of access terminals in the Internet of Things(Io T)environment has begun to grow explosively,and the computing power of many Io T devices can no longer meet the business needs of Io T applications.Edge computing and cloud computing models provide an effective way to solve this problem.The edge computing model and the cloud computing model have advantages in low-latency characteristics and adequacy of computing resources respectively.Therefore,studying the collaborative computing model of edge computing and cloud computing is of great significance for improving the computing capability of Io T terminal devices and the quality of Io T service applications.Combining the queuing model and the task transmission channel gain,the thesis studies the task offloading model and strategy in the edge-cloud collaborative computing environment with the main goal of reducing the completion latency of computing tasks,and mainly completes the following work:1.Aiming at the problem that the resource coordination is relatively simple in the task offloading of the edge-cloud collaborative computing,and the task waiting caused by a large number of concurrent tasks is not considered too much,the task offloading strategy in the edge-cloud collaborative environment is studied.This thesis constructs an average latency expectation model of multi-level task offloading based on queuing theory under the "device-edge-cloud" three-layer collaborative computing system model,which comprehensively considers the computing power and load of the device layer,edge node layer and cloud center.Thus,the objective function of average latency of tasks is further obtained.The Honey Badger Algorithm is used to solve the objective function,and the task is offloaded according to the optimal queuing system capacity of each end device and edge node layer.The results show that the task offloading strategy can effectively reduce the task completion latency in the edge-cloud collaborative computing environment when compared with loss-based queuing system offloading strategy,non-coordinated offloading strategy and random offloading strategy.2.The task offloading strategy under unstable wireless channel transmission rate is studied.Based on the experimental results of work 1 and the analysis of actual scenarios,it can be seen that the unstable transmission rate of wireless channel has a great impact on the offloading and average latency of tasks.Based on the study of Work 1,considering the wireless channel gain from each end device to edge node layer,the task latency expectation model under unstable wireless channel transmission is constructed,and the objective function is solved by the Optimized O-HBA algorithm.The task is offloaded according to the obtained optimal queuing system capacity of each end device and edge node layer.The experimental results show that the task offloading strategy has lower task completion latency than other task offloading strategies in unstable wireless channel environment.The research work shows that the task offloading strategy in the edge-cloud collaborative environment proposed in this thesis can effectively reduce the task completion latency,and also has a well performance in the environment where the wireless channel is unstable.