Research On User Allocation And Service Migration Optimization Method In Mobile Edge Computing Environment

Mobile edge computing(MEC)is proposed to give the network edge computing and storage capabilities,and its edge servers are placed close to users so that they can provide services to users in a very low-latency environment,making up for the shortcomings of the cloud computing model with high interaction latency,which can well meet the needs of some latency-sensitive applications for real-time computing and greatly The service experience of users is greatly improved.However,in the mobile edge computing environment,the edge server only has limited computing resources and storage capacity,and an unreasonable user allocation strategy will reduce the capacity of bearing users within the edge system and may cause resource waste;when users move to the signal range of the overloaded edge server,they may not be assigned to the edge server and wait in line or be assigned to the remote cloud server,for this reason,it is necessary to Reasonable user allocation algorithm is needed to enhance the user carrying capacity and resource utilization of the edge system.Second,as users move,when the network distance between users and their interactive edge servers is too large,the perceived service delay of users increases and service migration is required.However,due to the limitations of migration cost,privacy,and network congestion conditions,some service migration strategies will be difficult to coordinate many of these factors,and for this problem,service migration strategies need to be solved comprehensively in multiple target latitudes.To solve the above two problems,user allocation algorithm and seamless migration strategy are proposed in this thesis,respectively.The main contributions of this thesis are as follows:(1)We proposed an adaptive mobile path-aware user allocation algorithm,which first analyzes the user’s travel state,combines the road network data to predict the user’s future mobile trajectory,and proposes an allocation strategy based on this trajectory,using the expected dwell time of the user within the server signal range as the adaptation value to allocate the best edge server to the user and reduce its connection interruption due to being out of the server signal range;then An allocation decision adaptive adjustment algorithm is proposed based on the best adaptation algorithm,so as to further improve the user allocation ratio and the resource utilization of the server.(2)A location privacy-aware seamless service migration strategy based on NSGA-Ⅲ is proposed.The algorithm first encodes the migration strategy into samples in NSGA-Ⅲ,calculates the solution of each sample in the three metrics of user request service delay,location privacy,and migration cost,then iteratively selects the optimal migration strategy using NSGAⅢ,and then combines the Then,we weight each link in the two dimensions of communication delay and communication cost to form a new topology graph,and then use Dijkstra’s algorithm to select the set of migration paths,so as to reduce the service interruption time caused by service migration.(3)This thesis conducts a large number of experiments on open source road network dataset and user trajectory dataset.The experimental results show that the proposed method in this paper can effectively improve the service experience of mobile edge users and enhance the user carrying capacity,resource utilization and the degree of location privacy protection of users in the edge system,and reduce the cost of service migration.