Research On Task Offloading And Its Application In Ultra-Dense Edge Computing Networks

With the rapid development of 5G and information technology,terminal devices show a trend of explosive growth.Traditional network architecture is often unable to timely respond to such a large scale of device access and network capacity requirements due to the limitations of access capacity,spectrum resources and computing power.The ultra-dense network improves network capacity and access capability through the dense deployment of Small Base Stations(SBS).Mobile edge computing reduces application latency,improves terminal device battery life,and ensures high-performance and low-latency computing services by offloading complex computing tasks from terminals to servers connected to SBS.In order to alleviate the limitations of time delay,network capacity and computing resources during the task offloading of terminal devices,a new type of ultra-dense edge computing network framework is formed by combining the respective advantages of mobile edge computing and ultra-dense networks.However,the task offloading of ultra-dense edge computing networks still faces many challenges,such as the waste of communication resources due to the low utilization rate of wireless channel resources,the serious impact of spectrum interference on the quality of communication service,the offloading of mobile devices if they exceed the service range of the base station,and the change of channel state.The problem becomes more difficult.To this end,this paper will study the task offloading problem of ultra-dense edge computing networks from three aspects: energy consumption optimization,computing overhead optimization and smart grid user-side cost optimization:(1)For the task offloading problem oriented to energy consumption optimization,an energy consumption optimization model of ultra-dense edge computing network is designed,and the discussed optimization problem is further described in three aspects:task offloading,channel allocation,and power control.In the task offloading strategy,the state change of the wireless channel,the dynamic demand of the mobile device and the resource limitation of the server are mainly combined;In the channel allocation strategy,the interference limitation of spectrum resources is mainly considered;In the power control strategy,based on the golden section algorithm to find the optimal upload power.Finally,based on the Adaptive Genetic Algorithm Simulated Annealing(AGASA)algorithm,this paper obtains the optimal strategy for task offloading and channel assignment.Experimental results show that this strategy can minimize the energy consumption of task offloading while satisfying the deadline constraints.(2)For the task offloading problem oriented to computational cost optimization,a computational cost optimization model for ultra-dense edge computing networks is designed.In the task offloading strategy,the influence of the service range of the base station and the location information of the mobile device on the offloading result is mainly considered,and a task offloading scheme combining the device stay time and departure time is proposed.In the channel allocation strategy,a spectrum allocation scheme combining non-orthogonal Multiple Access(NOMA)transmission mode is proposed to minimize the channel gain difference.The multi-device power control algorithm obtains the optimal upload power.Finally,this paper obtains the optimal task offloading strategy and channel allocation strategy based on the Adaptive Genetics Algorithm Differential Evolution(AGADE)algorithm,and minimizes the computational overhead when the deadline constraint is met.The experimental results show that,compared with other schemes,the strategy proposed in this paper can reduce the computational overhead of the task offloading process well,and has good performance.(3)For the task offloading problem applied to the user side of the smart grid,a resource utilization cost optimization model for ultra-dense edge computing networks is designed.Due to the increasing number of terminal equipment access and the continuous advancement of grid intelligence,it is necessary to consider improving the grid capacity and the unloading efficiency of grid equipment requests.However,in the task offloading problem on the user side of the smart grid,most of the previous researches focus on the network architecture and algorithm level,while ignoring the impact of limited computing and communication resources on offloading strategies.In view of the above problems,this paper considers the influence of grid equipment,wireless channel and server resource price on the unloading results,and combines the types of grid equipment requests,and considers the interference of grid equipment as a constraint condition into the channel allocation strategy,and designs a method.A resource utilization cost optimization scheme for ultra-dense edge computing networks that satisfies deadlines and energy consumption constraints,and is used to solve the task offloading problem on the user side of smart grids.Among them,in order to obtain the optimal task offloading and channel allocation strategy,a Levy flight-MAPSO algorithm is proposed.The experimental results show that the impact of different prices on the cost of resource utilization is very significant,and the performance of the algorithm proposed in this paper is better than other algorithms.