Research On Resource Management Model And Algorithm In Cloud And Fog Network

With the popularity of large-scale mobile terminal devices,high-speed or ultra-low latency data traffic has exploded.The huge surge in traffic is not only due to the increasing user demands,but also to a greater extent due to the repeated transmission of popular content.Therefore,people have proposed a feasible solution based on the cloud and fog networks,which caches popular content in the mobile edge server in advance to reduce file transmission latency and improve user experience.In addition,since mobile devices are often equipped with low-capacity battery and limited-computation capability,the increasing applications cannot run efficiently on them.One of the possible solutions is to enable mobile devices to offload their intensive computation tasks to the mobile edge computing(MEC)servers to improve system performance.Based on this,how to realize the reasonable allocation of resources(such as storage resources,computation resources,bandwidth,power,etc.)has become a key issue that needs to be solved in cloud and fog networks.This thesis mainly studies the resource optimization management algorithm based on content caching and computation offloading in the cloud and fog networks.According to content caching,a joint cache placement and user-BS association scheme is studied;Based on computation offloading,the joint offloading decision-making and resource allocation strategy are explored.The specific research contents are as follows.Taking content caching into account,the thesis modeled the joint optimization problem of the content caching and user-BS association as a user efficiency maximization problem in Fog-RANs under the constraints of limited cache capacity and different user preferences,where user efficiency refers to the user utility obtained by spending a unit of time.Due to the NP-hard nature of optimization problems,traditional algorithms cannot solve this type of problem.Based on this,this thesis proposes a distributed sub-optimal solution to the caching and association problems.First,the original optimization problem is decomposed into the caching optimization sub-problem and the user-association sub-problem,and then the first sub-problem is solved by the maximization sub-module function method under the matroid constraint,and the second sub-problem is solved by the many-to-one matching algorithm.Finally,the convergence and complexity of the proposed algorithm are analyzed.Simulation results show that the algorithm can converge quickly,and the user efficiency is significantly improved compared with existing algorithms.According to computation offloading,based on the partial offloading model,the thesis first established a joint optimization model for computation offloading,bandwidth,and computation resource allocation with the goal of minimizing the average latency of users under the condition of limited system bandwidth and MEC server computation capacity.Then the Block Coordinate Descent(BCD)algorithm is used to decompose the strongly coupled joint optimization problem into two sub-problems,and finally a distributed joint optimization algorithm is proposed to solve it.The algorithm has linear computational complexity,can quickly converge to a sub-optimal solutions,and is suitable for network systems that require high real-time performance.The simulation results show that the proposed algorithm is significantly better than other comparison algorithms in terms of latency performance.On the basis of the above research,this thesis further jointly considers the energy consumption of tasks in the computation process.Based on the 0-1 offloading model,the offloading decision and resource allocation are jointly optimized to minimize the average computation overhead of users,where the computation overhead is a function related to computation time and energy consumption.Since the optimization problem is a mixed integer problem,this thesis decomposes it into the offloading decision sub-problem and the system bandwidth and computation resource allocation sub-problem,and obtains a feasible solution to the original problem by iteratively solving the two sub-problems.The simulation results show that the proposed algorithm can realize the adaptive adjustment of computation offloading and resource allocation,and greatly reduce the user's computation overhead under limited resource conditions.