Energy Efficient Edge Computing For Multi-task Min-max Fairness

With the development of technology,mobile devices have higher requirements for delay.Due to hardware limitations,mobile devices cannot provide enough electricity and computing resources required for delay-sensitive applications.Therefore,mobile edge computing is proposed to solve this problem.Most of existing works focus on how to utilize the computing performance of edge servers to reduce the energy consumption and execution time of mobile devices during task processing.However,the existing works have not yet a good solution to the problem of multi-device and multi-task energy balancing.Therefore,the aim of this thesis is to design an effective calculation model to solve the energy balancing problem of multidevice and multi-task.This thesis improves the existing works on the model of edge computing system for the energy balancing of multi-device and multi-task.In the model,this thesis adds a relay device node to use assisted transmission technology to accelerate the process of data transmission.At the same time,due to limited computing power,the relay device does not assist in the calculation of the task,but only plays a role in assisting the transmission.By assisting to transmit the tasks of mobile devices,the relay device can not only increase the transmission rate of the task,but also effectively relieve the pressure of data transmission in the network.In addition,because the relay device has the characteristics of strong mobility,it can effectively increase the communication coverage of the edge computing system,and improve the flexibility and real-time performance of the system.In order to solve the problem,this thesis designs a greedy algorithm for energy consumption balancing based on the Min-Max principle.The proposed algorithm can perform offloading and allocation of tasks for mobile devices.And it can minimize the difference in task energy consumption between mobile devices,while ensuring that the total energy consumption of mobile devices is as small as possible.Specifically,the proposed algorithm first calculates the total energy consumption of each subtask of each mobile device in turn,generates multiple energy consumption matrices,and thus generates an initial energy consumption allocation strategy.Then,the energy consumption allocation strategy is adjusted in turn until the final energy consumption allocation strategy is obtained.Finally,the proposed algorithm considers the energy allocation strategy of all subtasks,and judges whether its completion time meets its time constraints or not.If it is satisfied,a feasible energy consumption allocation strategy is obtained.In addition,this thesis analyzes the approximation ratio of the proposed algorithm.In order to evaluate the performance of the proposed algorithm,this thesis compares it with the total energy consumption optimization algorithm and the random algorithm.Experimental results show that,compared with the two baseline algorithms,the proposed algorithm can improve the performance of 66.59% in terms of energy consumption balancing,when the minimum transmission power of the mobile device is 5d Bm.And it can improve the performance of 61.87%,in terms of energy consumption balancing,when the minimum transmission power of the mobile device is 6d Bm,compared with the two baseline algorithms.Meanwhile,under the classic task topology,when the minimum transmission power of the mobile device is set to 5d Bm and 6d Bm,respectively,the proposed algorithm can almost obtain the performance similar to the brute force algorithm.