Research On The Long-Term Green Response Scheme For Multiple Applications In MEC Network

The continuous advancement of intelligent society has stimulated the rapid development of new Io T applications,including intelligent diagnosis and treatment and so on,which has put forward a high demand for the computing resource of intelligent terminals.However,due to the limited size of intelligent terminals and manufacturing process of central processing unit(CPU),limited computing resource of intelligent terminals is difficult to meet the high computing demands of a large number of new Io T applications.To solve this problem,multi-access edge computing is proposed,which aims to meet the extreme response requirements of new Io T applications.However,a large number of new Io T applications bring great pressure to edge cloud with limited computing resources.At the same time,the frequent information interaction among terminals and access nodes leads to a sharp increase in energy consumption of terminals and systems.On the one hand,it increases the energy consumption of the grid and greenhouse;On the other hand,the battery life of the intelligent terminal is greatly limited.Thus,it is necessary to study long-term green response for multiple applications.Firstly,this thesis fully explores the computing cooperation mechanism among terminals and reasonably adjusts the processing sequence of multi-application requests,which aims to achieve green processing of multiple application requests.At the same time,this thesis makes full use of renewable energy harvesting and SWIPT technology,which aims at realizing long-term green data transmission process.In general,this thesis is devoted to the study of green processing and long-term green transmission process for multiple application requests,which ensures the long-term green response of multiple application requests and improve the utilization of computing resources and renewable energy.The specific research content includes:1)Research on green offloading and processing scheme for multiple application requests: Considering a multi-application MEC system,by taking full advantage of the computing resource cooperation mechanism among terminals combined with the edge cloud offloading,the total power minimization problem of terminal devices is proposed which is constrained by the limited power,computing resource supply and application processing delay requirements of terminal devices.Based on greedy algorithm and bubble sorting method,the low-complexity Green Offloading and Application Processing Sorting Algorithm(GOAPSA)is proposed and designed which gives the joint optimization scheme of computing offloading decision,power allocation and application processing sequence.Besides,the complexity analysis of GOAPSA is given.Simulation results show that the system under the control of GOAPSA can effectively reduce the total energy consumption and increase the number of effective application requests.2)Research on green long-term power allocation and scheduling scheme: Considering a hybrid energy supply based SWIPT system,by designing a hybrid power supply model,a service-driven grid power consumption minimization problem is built which is constrained by the limited battery capacity of wireless access nodes,heterogeneous information acquisition and energy harvesting requirements of terminal devices.Based on Block Coordinate Descent(BCD)and the classification discussion method,Adaptive Power Scheduling and Splitting Algorithm(APSSA)is proposed which provides the joint optimization scheme of renewable energy cooperation,transmission power allocation and power splitting.Besides,the complexity analysis of GOAPSA is given.Simulation results show that the system under the control of our APSSA can achieve the long-term information acquisition requirements for all intelligent terminals,and reduce the grid power consumption by about 76.6%.