Research On Task Offloading And Resource Scheduling Strategy For Heterogeneous Networks In Fog Computing

As an emerging computing paradigm in recent years,fog computing has gradually become one of most popular technologies in the domain of Io T.On the one hand,compared with the computing locally by the terminal devices,the fog computing architecture can make use of the computing,storage and application resources which are deployed in the fog servers to effectively process most of information in real time.It expands the available resources of terminal devices and improves the mass data processing capability of communication networks.On the other hand,compared with the cloud computing,resources and service are deployed closer to users,which shortens the transmission distance of offloading tasks.It actually reduces the delay caused by transmission for task,then the network energy consumption is reduced too.Finally,the experience with Io T applications is improved and the network bottleneck problem of cloud computing is effectively solved.However,the domestic fog computing industry is still in the early stage of development at present.In this background,it is of great significance to study the fog computing framework and related issues to promote the industry development.In a practical sense,fog computing is not a substitute but a supplement to cloud computing.Distributed fog computing servers are far less resource-rich than cloud computing data centers,so cloud computing is still essential for some large-scale tasks.Therefore,how to realize the effective coupling of fog computing and cloud computing architecture and form an efficient coordination system to deal with the offloading tasks is a key problem.Then,towards the collaborative computing services,in order to achieve appropriate scheduling of offloading tasks,some effective strategies with arranging and dispatching need to be proposed.In this paper,two specific scenarios and network architectures for deploying fog computing services in practice are studied.And numerical indexes and calculation methods to measure the offloading strategy are developed.Based on the above,schemes are proposed for the formulation and optimization of offloading strategy.The main contribution of this paper is summarized as follows.Firstly,a fog computing service application scenario indoor including multiple users and multiple fog nodes is proposed.In the offloading process,the weight coefficient based on task requirements is used to balance the delay and energy consumption,then the sum of quantitative delay and energy consumption is used as offloading cost to formulate a mathematical model of optimization problem.Aiming at the optimization problem in the model,we apply an improved integer particle swarm optimization algorithm to solve it.The simulation results show that the proposed algorithm could have a better performance.Secondly,towards the offloading and computing requirements of outdoor mobile devices,we propose a cloud-fog collaborative offloading system model combining the latest ultra-dense wireless heterogeneous mobile communication network architecture.By applying iterative optimization,weighted max min fair algorithm and improved integer particle swarm optimization algorithm,the joint optimization scheme of offloading decision,wireless bandwidth resource allocation and transmission power optimization is realized.Simulation results show that the proposed scheme reduces the cost of the whole system and improves the efficiency of offloading.