Task Offloading Strategy For Location Privacy Protection

The rapid increase of mobile devices and big data has caused a large number of tasks to be concentrated on the mobile edge side.The complexity of task offloading and scheduling in mobile edge computing is increasing.Wrong offloading decisions not only make users' privacy easy to leak,but also bring huge losses to service providers.Therefore,task offloading based on privacy protection in mobile edge computing has become an important research point in the industry.This thesis studies the task offloading strategy for location privacy protection.The specific research content includes:(1)Combining the concept of virtual location space and linear programming algorithm,the location privacy protection strategy of edge wireless devices based on location ambiguity is studied;(2)Aiming at the problem of revenue and energy consumption in task offloading,in order to perform as many tasks as possible with limited resources,based on the research results(1),a task offloading strategy based on deep reinforcement learning is designed.The specific research results mainly included as follows:1.Aiming at the problem that device privacy is easy to leak in current mobile edge computing,combined with the concept of virtual location space and linear programming algorithm,a privacy protection strategy for edge wireless devices based on location ambiguity is proposed,which reduces transmission while considering location privacy protection delay.A large number of experiments have verified that this algorithm can reduce latency while protecting location privacy.When there are many surrounding servers and the joint inference of user privacy is not considered,compared with other algorithms,the security level is increased by about 7%.2.In order to increase the revenue of service providers and increase user utility,the task offloading strategy must perform as many tasks as possible under limited resources,and reasonable scheduling to improve resource utilization.Therefore,this thesis designs a task offloading algorithm based on deep reinforcement learning.First,extract the distance from the privacy-protected task offload data and allocate bandwidth.Then,a neural network is constructed to output the offloading decision.Experimental comparison proves that the strategy proposed in this thesis has obvious advantages in terms of privacy protection and saving energy and time.In view of the low energy consumption of tasks in mobile edge computing,compared with other algorithms,energy consumption and time savings have increased by about 28%.Finally,this thesis designs a system based on the research work done,which uses the Django framework in python to build a task offloading system for location privacy protection.The system displays the research step data,result data and basic parameters of the system and other information,which verifies the effectiveness of the strategy and algorithm in this article.