The increasing urgency and expansion of information systems implementing the Internet of Things (IoT) concept determine the importance of the investigation in the field of protection mechanisms against a wide range of information security threats. The increased complexity of such investigation is determined by a low structuring and formalization of expert knowledge on IoT systems. The paper encompasses an approach to elicitation and use of expert knowledge on detection of anomalies in data as well as their usage as an input for automated means aimed at monitoring security components of IoT.

The development of systems based on embedded components is a challenging task because of their distributed, reactive and real-time nature. From a security point of view, embedded devices are basically systems owned by a certain entity, used frequently as part of systems owned by other entities and operated in a potentially hostile environment. The development of security-enhanced systems of embedded components is a difficult task due to different types of threats that may affect such systems, and because the security in systems of embedded devices is currently added as an additional feature when the development is advanced, or avoided as a superfluous characteristic. We present in this paper a methodology for the analysis and modeling of threats and attacks for systems of embedded components. The Intruder Model allows us to describe possible actions a potential intruder can accomplish, depending on his/her capabilities, resources, etc. Using this information, we can define a Threat Model that will specify the threats and attacks that affect different security properties in specific domains.

The paper proposes an approach to detection of anomalous data from sensors in cyber-physical systems on an example of a water supply supply system. The approach is based on methods of machine learning and modeling of technical systems. The primary data for machine learning was obtained on the developed software/hardware prototype of the water supply system by using a number of microcontrollers, sensors and actuators. The experiments confirmed the feasibility of the proposed approach. Several machine learning methods from scikit-learn library of the Python programming language were tested. As a result of the experiments we identified a learning method and its parameters ensuring the highest accuracy of the abnormal situations recognition.

Комбинированная методика проектирования защищенных встроенных устройств на примере системы охраны периметра