Method for Detecting Abnormal Activity in a Group of Mobile Robots.

The range of attacks implemented on wireless networks is quite wide. To avoid or reduce the likelihood of an attack, it is necessary to use various defense mechanisms. Existing protection mechanisms are not always suitable for robotic systems and may not fully provide the necessary level of security. Thus, it is necessary to develop new ways of protection, which would be specific to groups of mobile robots. In this study, we propose an analysis of the following cyber parameters: the power consumption and the residual energy, as well as an in-depth traffic analysis to evaluate the effectiveness of the attack and identify abnormal network. We realized an analysis of the behavior of robotic systems under normal conditions and determined that, by their nature, robotic systems have a static and uniform behavior. We developed an experimental stand, and also conducted a theoretical analysis to confirm our assumptions. We found that some indicators of the components of the robotic system change statically; that is, there was a little deviation from the mean. Thus, we identified a set of metrics that allow us to determine how static the operation of the robotic system and its components is. Metrics allow us to evaluate parameters such as power consumption, and incoming/outgoing/redirected/dropped network packets. The results obtained are important for creating an integrated system for detecting anomalies in robotic systems. At the same time, the robotic node can analyze these parameters independently and make calculations that do not greatly affect performance. The main idea of this paper is to define a set of metrics characterizing the static behavior of a robotic system for the further development of an anomaly detection system.