Research In Optimization Of SCADA System In IoT Of Agriculture

Recently, IoT(Internet of Things) develops rapidly and has been widely applied in agriculture. With the expansion of the scale of IoT in agriculture and the increase of network traffic, the system of SCADA(Supervisory Control And Data Acquisition), which locates at the bottom of the application layer of IoT, mainly encounters three problems. First, SCADA is poor with high concurrency of data accessing of the collection terminal. Second, on account of heterogeneous data standard, it is difficult to achieve banlance between universality and efficiency on data format. Additionally, instead of rapidly responsing message and real-time processing data, SCADA is usually bad at dealing with the data flow. Aimed at enhancing the SCADA’s performance in concurrency and data processing, this article abstracted a universal feature model from the SCADA system, and studied optimization technique in data acquisition, processing and storage.This paper mainly completed the following work:1. Abstracted a universal feature model from the SCADA system in agriculture. Discussed the technology of improving the ability of data process on windows-based network communication model, the data process model which based on application layer protocols and algorithms of data flow process and the data storage process which based on connection pool and query optimization.2. For the problem of information locating, resource management, memory page locking and data packet process in IOCP(I/O Completion Port Model) model, this article optimized object pool which is made up of private-thread-based list and proposed a strategy of balanceing object resource in object pool which based on lock-free data structure, CAS mechanisms and mandates stolen. This finally solves the problem of low performance of thread pool in concurrent accessing to shared resources.3. For the chaos in message formats of data communication protocol caused by inconsistent data standards, this article proposed a new format based on JSON and TLV, which solves the problem of inefficiency in data transmission and resolution. It results in achieving better balance between the bandwidth/memory and response time. Meanwhile, according to the feature of data flow in IoT of agriculture and the requirement in data processing, this article proposed a new real-time algorithm to deal the data flow.4. Designed and builded a SCADA system for application-layer-oriented in IoT. This system verified the univercity of the model and solved the problem of large-scale concurrent data access and process.When faced with high-speed data requests, the practical application of the system showed that the optimization in network communication increaseed the concurrency of the system, and improved the efficiency of communication. The optimization of the message format improved the efficiency in data transmission and encapsulation/parse, while the data flow processing algorithm reduced the system’s memory usage and improved the efficiency of real-time update.