A Joint Information And Energy Cooperation Scheme For Edge Computing Networks

With the development of information and communication technology,the need to handle computationally intensive tasks is growing.The conflict between intensive tasks and scarce resources is becoming more and more pronounced due to the limited computing power and energy available in individual devices.An effective solution is to shunt computing tasks to nearby locations with more computing power and sufficient energy supply.Edge computing networks(ECNs)have emerged as a result and are attracting a lot of academic attention.ECNs are a new type of hybrid network architecture.Most of the current research on ECNs is only concerned with computational offloading or energy collaboration between nodes in the network,ignoring the connection between computational offloading and energy.On the other hand,when studying computational offloading or energy collaboration,existing studies usually tend to focus on the whole ECN,ignoring the consideration of the individual characteristics of the network nodes.In addition,existing studies model the cost of edge network systems in a single way,lacking a comprehensive consideration of the costs of computation and energy,making the overall cost not optimal.To address the above problems,considering that the computational offloading strategy and energy collaboration strategy are coupled with each other,this paper investigates the information-energy collaboration scheme of ECN to achieve the optimal scheduling of computational resources and energy.The main research contents include:(1)An information-energy resource collaboration scheme for two nodes of an edge computing network is investigated.In this novel resource collaboration model,an Edge Node(EN)can offload computational tasks to other ENs and/or cloud servers,and the EN can both store energy and share energy with other EN nodes.To enable the EN to have a stable power supply and complete the computational tasks in a given time,this paper models the information-energy collaboration problem as an optimization problem that minimizes the cost of cloud computing and the cost of grid power purchase.In order to find the optimal computational offloading and energy collaboration strategy,this paper first analyses the computational task offloading strategy under different energy conditions,and then analyses the energy collaboration strategy according to the offloading strategy.In order to obtain the optimal collaboration strategy,a low-complexity hybrid greedy iterative algorithm is proposed to solve the optimization problem.Finally,the simulation results demonstrate that the proposed scheme can effectively reduce the cost of the edge network and has better stability compared with other existing resource collaboration schemes.(2)A multi-node information-energy resource collaboration scheme for edge computing networks is investigated.Considering a typical ECN,i.e.Vehicle Edge Computing Network(VECN),one of the key problems of its edge node Road Side Unit(RSU)is how to handle computationally intensive tasks in a timely manner and utilise energy efficiently.To this end,the computational energy consumption of ENs within a multi-node edge computing network is first modelled.Then,the computational task transfer and energy transfer methods between ENs are analysed,and an information-collaboration model of the edge network is developed to model the information-energy collaboration problem as a total expenditure minimisation problem to reduce cloud offloading costs and make full use of the renewable energy in the VECN.Further,considering the selfishness and rationality of RSUs,personal gains are introduced into the optimisation problem.Based on this,a low-complexity two-tier coalition formation game algorithm is proposed,and it is theoretically shown that the algorithm can converge to a stable Nash equilibrium after finite iterations.Finally,simulation results show that the proposed scheme in this paper outperforms existing schemes by reducing the edge network overhead while improving individual gains.