Research On Resource Management In Mobile Edge Networks

The popularity of the Internet of Things(IoT)terminal devices in the 5th generation mobile communication system(5G)and the proliferation of new mobile applications,have put forward the demand for the current 5G mobile communication network,such as ultra-low latency,super-large bandwidth and massive connection.In order to meet the above challenges,mobile edge net-work has been proposed and received extensive attention from academia and industry.By providing cloud capabilities at the edge of network,including computing,storage and business service capabilities,mobile edge network can not only solve the connectivity problem of massive devices in 5G mobile com-munication network,but also relive the loading pressure of core network,mak-ing it one of the most potential technologies.However,resource management in mobile edge networks becomes more complex due to the limited resources of edge nodes,dynamic changes of network environment and diverse user de-mands,which greatly restricts the deployment of "cloud" capability.Efficient resource management technology can make use of network resources,realize excellent resource management and optimal task offloading,and contribute to the realization of large-scale and distributed deployment of "cloud" capability at the edge of the network.Therefore,the key technology of efficient resource management is worthy of further study.Considering computation task offloading,real-time changes of network environment and massive connection,resource management strategies are de-signed respectively in this paper to optimize the network overhead,reduce the user delay and improve the social welfare of edge nodes.The main innovative work of this paper is as follows:Firstly,a resource allocation strategy based on task offloading is designed to minimize the network overhead.We first introduce the MEC technology into the mobile edge network and propose a mobile edge network architecture based on MEC.Besides,under the limited communication and computing resources,considering both delay and energy consumption,the overhead minimization problem is proposed.What's more,considering the conflict of interests be-tween mobile devices(MDs)in modeling problem,the game theory is used to transform the original problem into a potential game problem,and a resource allocation algorithm based on task offloading is proposed.In this algorithm,the offloading decision function based on threshold structure is first designed to optimize the access selection of MDs.Then,the potential game strategy is used to realize the efficient distribution of network resources.Additionally,the convergence of the proposed algorithm is analyzed theoretically.Finally,simulations show that the proposed algorithm can reduce the network overhead.Secondly,a delay-aware online resource scheduling mechanism is pro-posed to minimize the network delay.First of all,a network delay minimiza-tion problem is proposed through real-time task offloading.Then,we design an online resource scheduling mechanism,which jointly considers the back-haul transmission strategy and transmission rate control parameters,so as to alleviate network load imbalance and congestion.Finally,simulations show that the proposed resource scheduling mechanism based on MEC can reduce the offloading delay of computation tasks.Thirdly,a distributed resource allocation scheme based on matching theory is proposed to maximize the network revenue.We first introduce the blockchain technology into the mobile edge network.Through the combination,the effi-cient transaction offloading of MDs is realized,which ensures the smooth de-ployment of the blockchain application in terminal devices.Then,we jointly consider the mining cost and reward of fog nodes,the problem of revenue max-imization is proposed.Moreover,a two-sided matching game with one-sided preference model based on network revenue is designed by using matching the-ory.Additionally,a distributed matching algorithm is proposed to realize the efficient matching between transactions and edge nodes and the complexity of the proposed algorithm is analyzed by mathematical theory.Finally,simula-tions show that the distributed matching algorithm can effectively improve the revenue of edge nodes,and also reduce the offloading delay of transactions.