CONCEPTUAL MODEL OF ARTIFICIAL INTELLIGENCE AGENTS FOR CONTROLLING MOBILE OBJECT COMMUNICATION SYSTEMS

The present article is concerned with implementing artificial intelligence agents in order to control mobile object communication systems. The fundamental premise is that artificial intelligence agents interact with their environment through semantic communications and possess semantics-based structures within a dedicated context of mobile network management. Semantics is discussed as the most general formalism for describing management as a process of observation and decisionmaking with an agent-based action. The novelty of the proposed approach is determined by the use of semantics to express the semantic content of control commands, which allows to use different ways of communication between agents with the help of a semantic network and to apply both large language models and voice control of a mobile object with the generation of new commands on the basis of machine learning in future. The present study hypothesizes that a semantic and context-based management approach is optimal for artificial intelligence agents in applied mobile telecommunications management scenarios. The conceptual model of artificial intelligence agents for management that emerges from this hypothesis encompasses roles within the context of management tasks and interfaces for context-dependent interaction with the knowledge management system or other agents. The conceptual model of artificial intelligence agents for management is defined, characterized by the use of a two-level hierarchical model of agents with different system roles in the context of management tasks, taking into account the interaction with the knowledge management system within the framework of the scenario. Machine learning methods are regarded as external to artificial intelligence agents. The classification of artificial intelligence agents for mobile telecommunication systems management encompasses reflection agents, self-learning agents, and goal-oriented agents. The special section includes an example of the agents’ queries for the applied scenario of telecommunication management knowledge and format of resulting information.