Cost-efficient And QoS-aware Task Scheduling Algorithm In Fog Computing

With the rapid development of the Internet of Things,the traditional single cloud computing model has been increasingly unable to meet the needs of users for low computing latency.Therefore,based on cloud computing,scholars have developed the concept of fog computing.The nodes in fog computing have high mobility and distribution.In order to guarantee good user experience,computing platform runs under constant QoS maximization scheduling policy.How to reduce unnecessary costs caused by the operation of this strategy is worth studying The problem.Therefore,this thesis proposes a dynamic adaptive task scheduling method based on cost efficiency optimization and a credibility rating mechanism of fog nodes.The main research work is as follows:1.Aiming at the problem that the cost model construction factors in the current fog calculation are only the cost of renting fog nodes or the energy consumption cost,this thesis constructs a more accurate cost model for fog calculation.The model takes additional factors such as bandwidth cost and task gain into account,which can measure the actual cost of the system more accurately.2.Design a dynamic adaptive task scheduling method for fog computing-oriented cost efficiency optimization.The adaptive method can dynamically select a scheduling strategy,and reduce unnecessary cost consumption caused by the service provider's operation of maximizing the scheduling strategy with a constant QoS.This strategy selects the task processing delay and throughput to calculate the QoS violation rate each time a new task is executed,and divides the QoS into three states based on the QoS violation rate: imperceptible state,tolerable state,and unavailable status.Then according to the dynamically acquired system QoS status,the adaptive scheduling strategy corresponding to its status is selected to complete the scheduling.After experimentation,this strategy can accurately reduce the cost of fog computing platform under the premise of guaranteeing system QoS.3.As the nodes of fog computing are usually distributed at the edge of the network and have high mobility,it is easy to cause data loss and task interruption,as well as additional costs and delay.Based on the second point,the credibility rating mechanism of fog nodes is introduced.This mechanism evaluates the stability of the fog node according to its historical task operation,and gives priority to placing the tasks on the nodes with higher rating.The experiment proves that the introduction of this mechanism can effectively reduce the QoS violation rate.The code of the IFog Sim simulation platform was rewritten,experimental results show that the proposed strategy is optimized by 42.1%,35.5%,7.7%,28.1% and 1.5%respectively when compared with the original task priority scheduling strategy,random scheduling strategy,maximized QoS scheduling strategy,minimized cost scheduling strategy and FSPP(Fog Service Placement Problem)strategy.The experimental results show that the proposed scheduling algorithm can accurately reduce the cost on the premise of guaranteeing the QoS operation of the system,which proves the effectiveness and superiority of the algorithm...