Comparing Agent-based Control Architectures For Next Generation Telecommunication Network Infrastructures

• Published in: IFAC-PapersOnLine Vol. 53; no. 2; pp. 11062 - 11067
• Main Authors: Hernández, Marco Pérez, McFarlane, Duncan, Herrera, Manuel, Jain, Amit Kumar, Parlikad, Ajith Kumar
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 2020
• Subjects: control of large-scale network systems; Decentralized; distributed control; Multiagent systems; Optimization; Multiagent systems; Decentralized; distributed control; Optimization; control of large-scale network systems
• ISSN: 2405-8963; 2405-8963
• DOI: 10.1016/j.ifacol.2020.12.238

Multi-agent systems have been an effective choice for designing control systems that are flexible and agile. However, few attention has been given to the evaluation of the architectures of such systems. This becomes critical with the emerging requirements in complex domains such as digital network infrastructures. In this paper, we propose an approach for the evaluation of agent-based control architectures and introduce three multi-agent based architectures for the supervisory control of network service operations of the next generation of digital infrastructures. With the proposed approach, we evaluated the architectures and the implemented control systems prototypes under a realistic network infrastructure environment. Our approach has been effective to evaluate the candidate architectures. The results of communication overhead and reaction time, have shown that agent-based hierarchical and heterarchical-ring architectures have outperformed the heterarchical-complete network architecture.