Research And Implementation Of Sensor Access System Based On Stream Computing Model

With the rapid development of Internet of Things(IoT)technology,a large variety of sensors are applied to more and more scenes.Sensors are heterogeneous and differ in communication protocols,data formats and other aspects.In addition,as the demand for IoT services grows and the scale of applications increases,the number of sensors that need to be accessed in the IoT system is also getting larger and larger.The traditional single-machine access system is not sufficient for large-scale sensor access.Therefore,how to quickly access large-scale and multi-type sensors,and realize distributed sensing data processing is an urgent problem to be solved in the development of the IoT.At the same time,the sensing data reported by different sensors have different degrees of importance.The more important and urgent data,the higher the requirements for real-time performance.How to ensure that important data is prioritized is also a point of our concern.In order to solve the above problems,we have researched and realized the sensor access system based on streaming computing mode,which provides fast access and management for heterogeneous sensors,and fast real-time processing ability for large-scale sensing data.The system dynamically generates sensor adapters based on the heterogeneous parameters that the user submits for the sensors.These adapters run distributedly in the cluster built by stream processing engine,and realize the access and parsing of sensor data.At the same time,the parsed data is published to the distributed message system,and the upper layer applications of the IoT can obtain the sensing data by subscribing to the sensor theme.In order to prevent the loss of burst sensing data and improve the fault tolerance of the system,this paper designs a caching strategy of sensing data,and dynamically caches sensing data according to the task execution metrics.In addition,in order to ensure the real-time performance of important data to the maximum extent,this paper designs a priority-based task scheduling scheme.This scheme starts from three aspects:task assignment,resource allocation and resource reclamation.Through the preferential treatment of high-priority tasks,the execution performance of high-priority tasks and the processing speed of high-priority data are greatly enhanced.This paper introduces the sensor access system based on streaming computing mode mainly from three aspects.Firstly,it introduces the research background and significance of the sensor access system based on streaming computing mode and puts forward the problems to be solved;then,according to the problem,the functional analysis of the access system based on the streaming computing mode is carried out,and various functional modules of the system are designed and implemented;finally,the functional testing of each module of the system was carried out,and the overall performance of the system was also tested.