The Development Of Edge Intelligent Gateway Based On Smart Lamp Poles

Smart lamp pole is gradually becoming one of the best carriers to promote the rapid development of smart city.It provides intelligent services for modern society by deploying the "end,pipe,cloud" IoT system architecture.The sensing layer of the multi-functional smart lamp pole contains multiple and different intelligent modules and sensing nodes.The gateway can realize the interaction between the sensing layer and the cloud platform through protocol conversion and other means.However,in the current smart lamp pole IoT system centered on the cloud platform,the gateway only completes data forwarding and protocol conversion;and cloud platform completes data analysis,calculation and network management.It will inevitably lead to the shortcomings of too centralized computing resources,serious data transmission delay,network congestion and so on.Edge computing which is complementary to cloud computing,can make up for the above shortcomings by building a platform with intelligent computing and data management capabilities on the edge side.However,the edge computing nodes often have the problems of single basic resource layer,lack of virtualization layer and lack of edge virtual function.There are still some deficiencies to maximize the use of hardware resources,improve the flexibility of the system and complete the functional business isolation.Therefore,with the support of government sponsored research,this research has developed an edge intelligent gateway based on smart lamp poles,which realizes functional business isolation through hardware virtualization,supports multi protocol conversion,management and local edge computing,and can complete real-time processing and computing work such as detection,calculation and decision-making at the edge of the network.The local calculation is to determine the appropriate lighting parameters which include brightness and color temperature,according to a variety of real-time environmental parameters.The brightness adjustment technology is very mature,but the color temperature adjustment of street lamps is still in the stage of classification,and coordinated adjustment of color temperature and brightness for multiple environmental parameter adjustments is also in the initiation stage.The main content of the paper is as follows:(1)The design of the edge intelligent gateway.The smart lamp poles IoT system based on edge computing is designed,and the three layer deployment model of cloud and edge collaboration is adopted to deploy intelligent assets and intelligent gateway on the edge intelligent gateway;the functional framework and system structure of the edge intelligent gateway are designed,and the edge virtual tasks,such as bus protocol adaptation,asset management and optimized computing,are realized on the basis of realizing the hardware virtualization of the edge intelligent gateway.(2)The design and implementation of embedded platform virtualization for edge intelligent gateway.Briefly introduces the basis of virtualization technology,analyzes the differences of different embedded microprocessor architectures,selects the virtualization technology suitable for embedded microprocessor,determines the mode of deployment multi virtual machine for multi business,and completes the Xen embedded virtualization design of intelligent gateway.The embedded microprocessor motherboards of ARMv8 architecture and ARMv7 architecture were selected to complete the Xen virtualization independently.The experimental verification proved the successful construction of the Xen virtualization platform.Based on the analysis results,the edge intelligence gateway chooses the motherboard of ARMv7 architecture as the development platform for subsequent lightweight applications.(3)The development of the intelligent gateway and intelligent asset.At the intelligent gateway layer,the network topology is designed,protocols for Ethernet,ZigBee,and RS485 communications supported by the edge intelligent gateway are designed,and the multi-protocol conversion from Ethernet to ZigBee / RS485 messages is completed.For intelligent assets,the idea of communication layer is proposed to realize the collection of real-time detection data,design data management framework and storage mapping table,and design configuration process for intelligent gateway and sensor node.Finally,an experimental platform is established to verify the functions of the edge intelligent gateway for multiple communications,protocol conversion,real-time data collection,system configuration,and network health diagnosis.(4)Edge calculation of color temperature adjustment parameters.On the basis of realizing the Xen virtualization of edge intelligent gateway,a DomU is created for the edge temperature business of color temperature adjustment parameters,and the inter-domain communication model of the edge intelligent gateway Xen is designed.Then according to the four real-time environmental parameters of temperature,humidity,light intensity and PM2.5 concentration,a multi-input and dual-output fuzzy computing system is designed to calculate the current optimal brightness and color temperature of the current light based on the multi-environment parameters.Through the simulation of the system model and the test of the actual developed application program,the results show that the designed multi-parameter fuzzy computing system can complete the expected street light parameter calculation task on the edge of the network,with better real-time performance.