Design And Implementation Of Smart Street Lamp Centralized Controller Based On Edge Computing

With the development of smart city,street lamps,as an important infrastructure of the city,have access to a large number of sensors and integrated functions such as road lighting and environmental monitoring.Smart street lamps have emerged at the historic moment.Smart street lamps generate massive amounts of data at the access layer,and the need for a centralized controller on the edge of the smart street lamp network to analyze and process data is increasing.How to effectively improve the data analysis and processing capabilities of the centralized controller has become an urgent problem to be solved in the development of smart street lamps.To this end,a smart street lamp centralized controller based on Edge Computing(EC)is designed in this thesis.It has edge processing functions and EC network task matching functions,which can reduce the data transmission volume of the smart street lamp network and improve the task execution efficiency of the network edge side,thereby alleviating the pressure of network transmission and cloud server processing massive data.The main contents of this thesis are as follows:1.According to the research status of the smart street lamp,centralized controller and EC at home and abroad,the overall scheme of smart street lamp centralized controller is designed.2.The edge processing function of the smart street lamp centralized controller is designed,so local analysis of power data and environmental parameters and image processing are implemented.The designed centralized controller improves the problem that massive data transmission occupies a lot of bandwidth and cloud server resources,and network transmission delays cause abnormal monitoring to be untimely and unable to control network edge devices in real time.3.A Firework-based Task and Resource Matching(FTRM)mechanism for EC networks is proposed,which improves the problem that tasks with greater resource requirements are executed inefficiently or even not when the centralized controller has insufficient resources.Experimental results show that the proposed mechanism can effectively distribute tasks to multiple nodes for coordinated execution,and can significantly shorten the task execution time.4.The smart street lamp centralized controller hardware and software areimplemented and tested.The test results show that the smart street lamp centralized controller can collect and send power data,environmental parameters and pictures spontaneously or according to the server instructions.The edge processing function of the smart street lamp centralized controller cooperates with the cloud to analyze power data and environmental parameters and identify license plates,which greatly reduces the amount of data uploaded to the cloud.