Research On Task Offloading Strategy Of Ultra Dense Networks Based On Mobile Edge Computing

With the continuous upgrade of mobile networks and the explosive growth of the number of mobile Internet users,the fifth generation mobile communication(5G)is facing new challenges of explosive data traffic growth and massive device connectivity.However,limited by the limited resources of mobile devices,it will not be able to fully meet the needs of users for certain services.Mobile Edge Computing(MEC)has abundant computing resources and features close to users,which will effectively improve the quality of service of users.At the same time,ultra-dense network(UDN)can enhance coverage and reduce end-to-end delay.In the future 5G network,the network architecture of the co-existence of UDN and MEC will meet users' service needs for high-speed,large-connection,low latency,low power consumption and so on.However,MEC's limited computing resources and complex network environments in ultra-dense networks make resource allocation,interference management and load imbalances a key factor that constrains MEC scalability and network performance.Therefore,this thesis mainly considers the limited resources and imbalanced network load,and studies the task offloading strategy of ultra-dense networks based on mobile edge computing.The main contributions are as follows:Considering the effect of different delay requirements and resource allocation on the performance of task offloading in ultra-dense networks,a joint optimization problem of offloading decision and resource allocation is formulated to optimize the total energy consumption of users,and a centralized optimization scheme is proposed to solve this problem.The coordinate descent method is first used to optimize the user's offloading decision.At the same time,considering the user's delay constraints,the improved Hungarian algorithm and the greedy algorithm are used to perform user-level sub-channel allocation under a specific offloading decision.Then,we turn the problem of minimizing energy consumption into a power optimization problem,and obtain the optimal transmission power of the users through the convex optimization method.The simulation results show that the proposed offloading scheme can save more energy than other schemes and effectively improve the performance of the system.Considering the problem of congestion delay caused by unbalanced load of MEC due to uneven distribution of users and MEC servers,a task offloading strategy based on business load balancing is proposed to optimize the system delay consumption.Depending on the traffic load,the user can offload some tasks to the appropriate MEC based on partial task offloading mode to minimize the system delay.Firstly,an iterative algorithm is used to optimize the task's offloading decision,and then the user's optimal offloading rate is found by the interior point method according to the user's offloading decision.Finally,the optimal task's offloading rate and offloading decision are found by iteration of the algorithm.The simulation results show that the proposed offloading strategy effectively achieves load balancing and optimizes the performance of the system.