Research On Computing Offloading Technology For IoT Edge Gateway Scenarios

With the rapid development of the intelligent Internet of Things,computing-intensive and latency-sensitive applications running on terminal devices are gradually prospering.The traditional cloud computing paradigm often brings high transmission costs and unacceptable processing delays,which cannot meet the low-latency and high-bandwidth requirements of terminal device applications.Mobile edge computing has become a hotspot of research and application.The edge gateway is one of the core devices in the mobile edge computing scenario of the Internet of Things.By partially or completely migrating the application tasks of the terminal device with limited resources to the edge gateway at the entrance of the wireless access network,it can make up for the computing,storage,Insufficient energy consumption,etc.Therefore,it is necessary to design a reasonable and reliable computing offloading strategy according to the actual IoT scenarios.Therefore,this paper mainly studies the computing offloading technology for IoT edge gateway scenarios.The main tasks of this paper are as follows:1.A computing offloading strategy for single-gateway IoT mobile edge computing scenarios is proposed.Aiming at this scenario,this paper designs an intelligent,dynamic and reliable computing offloading strategy for decision-making terminal task offloading and gateway resource allocation.By establishing the terminal task model,communication model,delay and energy consumption model,and integrating the three evaluation indicators of delay,energy consumption and task priority,the system objective optimization function is designed to minimize the joint cost of system delay and energy consumption.The optimization process is modeled as a Markov decision process,and a deep reinforcement learning algorithm is used to solve the computational offloading strategy in the dynamic environment of the system.The experimental results show that the computational offloading strategy proposed in this paper can minimize the joint execution cost of terminal task delay and energy consumption,and significantly improve the performance of various systems in this scenario.2.A computing offloading strategy for multi-gateway collaborative IoT mobile edge computing scenarios is proposed.This paper extends the single-gateway IoT MEC scenario to multi-gateway IoT MEC,and conducts in-depth research on the computing offloading technology in the multi-gateway MEC scenario.Firstly,this paper designs a gateway selection and allocation algorithm for dynamic decision-making,considering the dynamically changing network status and equipment load in the area,taking the long-term operation cost minimization of multi-gateway system as the decisionmaking optimization goal,and introducing the load balancing penalty on the gateway side to form Multiple on-premises edge networks with gateways at their core.Secondly,this paper designs a multi-gateway unloading collaboration algorithm based on asynchronous greedy strategy.The optimization goal is to minimize the cost of executing tasks on the gateway side,establish a cooperative relationship between gateways,and use idle gateway resources to solve the computational overload problem caused by the limited resources of a single gateway.Finally,the high-performance computing offloading strategy of multi-gateway IoT MEC system is formulated by combining single-gateway computing offloading strategy,dynamic gateway selection and allocation algorithm,and multi-gateway offloading collaboration algorithm.The experimental results show that the algorithm and strategy proposed in this paper can further improve the overall performance of the system in this scenario,and at the same time reduce the long-term operating cost of the system,and ensure the load balance among multiple gateways.In the IoT mobile edge computing scenario,the computing offloading technology based on the IoT edge gateway proposed in this paper can significantly improve the performance of the IoT mobile edge computing system,meet different IoT application scenarios,and provide a feasible research idea for the current development of the smart IoT.