基于边缘计算和稳态遗传算法的AIoT资源调度研究

Studies on Resource Scheduling for AIoT Using Edge Computing and Steady State Genetic Algorithm

在万物互联时代,智能物联网(Artificial Intelligence of Things,AIoT)是人工智能与物联网融合发展的新兴方向.在面对边缘海量数据和连接时,如何进行资源负载的调度是AIoT目前亟待解决的关键问题.边缘计算(Edge Computing)是对云计算的补充和发展,具有低时延、位置敏感和无线接入等优点,能在网络边缘进行高效部署.该文提出了一种基于边缘计算和遗传算法的AIoT资源调度方法.首先,基于边缘计算构建了云边融合的AIoT分层网络资源管理架构.然后,对其资源调度问题进行了数学建模并使用稳态分组遗传算法(SSGGA)进行了优化.最后,根据得到的优化方案制定了计算资源调度策略.另外,基于iFogSim平台搭建了实验环境,仿真结果验证了该文所设计的资源调度策略可有效降低高负载下AIoT各节点的处理时延,并有效提升网络内设备的能量利用效率.该文的工作对于推动AIoT的发展具有良好的理论和应用价值.

Artificial Intelligence of Things(AIoT)is an emerging direction for the integration and development of artificial intelligence and the Internet of Things in the era of Internet of Everything.At present,how to schedule resource loads in the face of massive data and connections at the edge is a key issue that AIoT needs to solve urgently.Edge computing is a supplement and development to cloud computing.It has the advantages of low latency,location sensitivity,and wireless access,and can be efficiently deployed at the edge of the network.In this paper,we propose an AIoT resource scheduling method based on edge computing and genetic algorithm.First of all,we build a cloud-edge integrated AIoT hierarchical network resource management architecture based on edge computing.Then we mathematically modeled the resource scheduling problem and optimize it using SSGGA genetic algorithm.Finally,we formulate a computing resource scheduling strategy according to the obtained optimization scheme.In addition,we set up an experimental environment on the iFogSim platform and the simulation results verify that the resource scheduling strategy designed in this paper can effectively reduce the processing delay of each AIoT node under high load,and effectively improve the energy utilization efficiency of the equipment in the network.The work of this paper has good theoretical and application value for promoting the development of AIoT.