Research On Energy Consumption Optimization Of Joint Task Offload And Resource Management In Edge Computing

With the continuous and rapid development of the Internet of Things and 5G mobile communication technology,the amount of network data is exploding,posing a huge challenge to the performance of intelligent devices.Limited computing resources are difficult to meet data processing requirements.As a new computing mode,edge computing can effectively solve the current problem,but the limited battery capacity of intelligent devices limits the application scope and effect of edge computing,and the short standby time of devices will also seriously affect the user experience.Ensuring the endurance of the device and reducing the energy consumption of the device are important research contents in edge computing at present.In this paper,energy consumption optimization of intelligent devices in edge computing is studied in depth,including energy consumption optimization through task unloading strategy and resource allocation in different scenarios of intensive tasks and delay sensitive tasks.The main tasks are as follows:(1)The research background and significance are described,and the current research status of energy consumption optimization in edge computing at home and abroad is analyzed.This article focuses on the characteristics of intensive and delay sensitive computing tasks,taking into account system queue length and task delay constraints,and proposes a multi user and multi server model and an edge cloud collaborative computing model.The model is mathematically modeled and analyzed.(2)This article proposes an energy optimization algorithm based on queue stability for scenarios of intensive computing tasks.The algorithm is based on Lyapunov optimization and takes into account the queue length and computing resources of the current time slot for each user to unload tasks.Under the premise of ensuring queue stability,the sub problem is solved to obtain transmission power and computing frequency within each time slot,By setting step size adjustment parameters,the minimum energy consumption of users under different weights is obtained.Experiments have shown that this algorithm can reduce users’ energy consumption by approximately 39%on the original basis.Changing the computing mode to edge cloud collaborative computing can reduce users’ energy consumption by approximately 55%.(3)For the scenario of delay sensitive computing tasks,this paper proposes a Lagrange iterative optimization algorithm considering the delay constraints of tasks.Through the phase analysis and modeling of task unloading,the minimum energy consumption optimization problem of joint local computing ratio,transmission power and computing resources is obtained.The optimal solution of this non Convex function is obtained by using KKT conditions and iterative calculation of local computing ratio.For delay sensitive tasks,delay constraints need to be considered.Two mechanisms are proposed to enhance the algorithm’s ability to obtain global optimal solutions,ensuring that more tasks are unloaded under delay constraints at lower energy consumption levels.By comparing the simulation results with the two heuristic algorithms,the improved algorithm reduces device energy consumption by 15.5%while reducing maximum computational complexity,and significantly reduces program runtime.