Research And Implementation Of Iot Data Acquisition System Based On Cloud-edge Collaboration

IoT data acquisition is one of the important foundations supporting IoT applications.With the continuous development of computer technology,there are higher requirements for the delay,availability and concurrency of IoT data acquisition.In this paper,we study and develop an IoT data acquisition system based on cloud-edge collaboration.The main contributions are as follows:(1)Real-time acquisition of sensing data based on cloud-edge collaborationAs a new deployment solution,cloud-edge collaborative computing can provide low latency computing power support for IoT sensing data acquisition services.In this paper,we design a system architecture of "endedge-cloud" collaborative computing.By dividing the tasks of IoT sensing data acquisition and rationally designing the functions of end,edge and cloud,the delay of data acquisition tasks can be effectively reduced.During the data aggregation operation,through modeling and analysis of edge nodes,the minimum delay problem is transformed into a minimum Steiner tree solution problem,and the optimal edge node is selected to offload the task,provide data aggregation services,and ensure that the data acquisition service time minimize delay.(2)High availability load balancing strategyThe system automatically deploys data acquisition containerized applications to the service cluster through Kubernetes.Through the monitoring of service status,Kubernetes performs health checks on the server cluster and performs self-healing or state migration when nodes run abnormally,ensuring the stability of the data acquisition system.In addition,in this paper,we design a load balancing strategy based on Kubernetes,which comprehensively considers the CPU,memory,disk IO and other factors of the data acquistion service node,quantifies the load balancing capability of the service node,and selects the optimal node that provides the data acquisition service after switching according to the calculation result.The problem of resource scheduling improves the availability of the system to a certain extent.(3)Implementation of IoT data acquisition systemThe IoT sensing data has the characteristics of multi-source and heterogeneous.In this paper,we divide the IoT data acquisition system into three modules:protocol management,data conversion and data preprocessing and rationally design its functions.It can realize the dynamic access of IoT entity resources and enable concurrently realize data acquisition of IoT sensing data,protocol analysis and data processing operations,which provide data support for subsequent IoT upper-layer applications.In addition,the IoT data acquisition system can also encapsulate data and protocols after receiving remote control commands from upper-layer applications,and send them to the IoT entity resources to realize the control function of the IoT entity devices.