Design And Implementation Of Edge Gateway System Based On Linux

With the rapid development of the Internet of Things and mobile Internet services,blockchain,big data,and artificial intelligence technologies are also developing rapidly.The ever-increasing demand for Io T applications has higher requirements for latency,capacity,and security.However,the traditional cloud computing architecture is not flexible enough,and only processing services through cloud servers often brings high latency and network congestion,resulting in often limited applications in the Internet of Things.Therefore,it is particularly important to reduce time delay and other factors by deploying edge nodes.In order to enable edge computing to play a better role in the Internet of Things,an edge gateway system with edge computing features has been designed and developed.Based on the research of wireless communication protocols,Linux systems,routing algorithms,etc.,this thesis designs and implements the functions of edge gateways.The main work content is as follows:(1)Design and implement the multi-link selection algorithm of the edge gateway system.The northbound edge gateway performs data transmission between heterogeneous networks and cloud servers.When a link is of poor communication quality or a link is disconnected,it usually affects data transmission and even causes data loss.Many designs are designed for this phenomenon.The link selection algorithm evaluates the current link status in time,and considers the situation of network fluctuations.When evaluating the current link quality,a forgetting factor is added to comprehensively consider the current link quality and the link quality in the previous period of time..When the communication quality of the current link is found to be poor,the remaining links can be comprehensively evaluated and the optimal link can be selected for handover.Experimental results show that the algorithm can evaluate the current link quality in real time and execute a handover strategy based on changes in link communication quality.(2)Design and implement a business-driven routing algorithm for multi-edge gateways.Different types of services often have different policy requirements.For example,audio and video services are more sensitive to bandwidth,but not particularly sensitive to delay,and signaling services are more sensitive to delay.In response to this phenomenon,a business-driven routing algorithm is proposed.Different services are distinguished by labeling different services in the PREROUTING phase.In the POSTROUTING phase,the corresponding routing strategy will be executed according to the label matching.Signaling services implement the shortest path algorithm,while streaming media services switch to use the maximum average path algorithm.Experimental results show that the business-driven routing algorithm can switch for different business types,and can improve the processing efficiency of different businesses.(3)Design and implement an edge gateway system.First,the software and hardware of the edge gateway system are designed,and secondly,the signaling scheduling function of the edge gateway system is designed for the situation that the terminal data cannot be transmitted due to the node damage or loss in some scenarios such as smart fire protection.Detect the online status of the subsystem by sending heartbeat packets with the subsystems,comprehensively evaluate the remaining online subsystems when the subsystem is offline or damaged,and select the optimal subsystem to connect to the terminals of the lost subsystem to ensure terminal data The stable transmission can achieve the characteristics of collaborative intelligence.Experimental results prove that when the subsystem is disconnected or damaged,the connected terminals can be dispatched to the appropriate subsystem to continue working,ensuring that the transmitted data will not be lost.