Software tools communication process models for Modbus TCP/RTU for diagnostics using machine learning approaches. Issue 1 (March 2021)

• Record Type: Journal Article
• Authors: Sosnovskiy, Yuri; Lapina, Maria; Lapin, Vitalii; Mecella, Massimo
• Is Part Of: IOP conference series. Volume 1069:Issue 1(2021)
• Publication Date: 2021-03
• Subjects: Software tools -- Machine learning -- Communication process models -- Generating data sets; Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/1069/1/012033
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: The work proposes a software solution to the problem of generating data sets that simulate information control processes in microprocessor control systems and automated process control system (APCS) at the lower and middle levels using the Modbus TCP/RTU protocol. Such datasets (dumps) with controlled characteristics, records of computer attacks and information and technical impact (ITI) of various types are advisable to use for training spacecraft recognition systems based on artificial intelligence and machine learning. Also, due to the possibility of introducing records of any types of attacks, it is possible to test intrusion detection systems, intelligent sensors of the spacecraft for their efficiency. The analysis of the existing software, the possibility of its combination to solve the problem is given. The structure of a data transmission system (DTS) fragment of a real VCE ACS and its representation as a set of device identifiers, addresses, registers and data types. The model of communication processes Modbus TCP/RTU was built, the results of modeling are presented. The features and limitations of the software used are discussed, as well as the prospects for improving the model.