The growing popularity of the use of group robotics, including swarm robotic systems (SRS), actualizes the issues of information security. Known approaches to detecting malicious behavior of agents or malicious information do not take into account the scalability and decentralization properties of SRS, which does not allow ensuring the integrity of information circulating through communication channels within SRS. In turn, the dissemination of malicious information in the process of distributing tasks between SRS agents initially reduces the efficiency of performing these tasks, that is, an attack is carried out on the very first and most critical stage of the system's functioning. The purpose of this work is to improve the efficiency of the functioning of SRS agents in the presence of malicious agents by developing a method for detecting and counteracting the spread of malicious information. The elements of scientific novelty of this work include the following. As part of solving the problem, a number of specific criteria are proposed that take into account the distribution of tasks in the SRS, as well as a classifier based on an artificial neural network to detect malicious information. To improve the accuracy of detection and counteracting the spread of malicious information in SRS, a modification of the reputation mechanism is proposed. A distinctive feature of the modification is not only the formation of an indicator of the truth of the message information in the process of task distribution, but also the assessment of the influence of malicious agents on the process of forming this indicator. The presented solution is implemented in the form of software in the Python programming language, which can be used in modeling decentralized control systems of SRS.

Keywords: swarm robotic systems, task distribution, artificial neural networks, trust and reputation mechanism