Nigerian Journal of Technological Development (Mar 2024)
Consensus Issues in Multi-Agent-Based Distributed Control with Communication Link Impairments
Abstract
In multi-agent systems, achieving consensus among autonomous agents is a fundamental problem with wide-ranging applications, from autonomous robotics to distributed sensor networks. However, the real-world deployment of such systems often involves communication links prone to impairments, including packet loss, delays, and network congestion. These communication challenges present formidable obstacles to achieving consensus reliably and efficiently. In this paper, consensus protocols were introduced for network with and without communication impairments and convergence analysis were provided in all the cases. The intricate dynamics of consensus issues in multi-agent-based distributed control under the influence of communication link impairments, connectivity and consensus protocol were established. Undirected communication graphs used to model the topology for agents’ connectivity is significant to addressing consensus issues of communicating agents. The paper also discusses the tradeoffs and design considerations in developing consensus strategies resilient to communication failures while optimizing performance. Simulation results show that an isolated agent in a network can achieve consensus only when there is a reference value. It was also established that communication impairments significantly degrade the performance of distributed agents in a network.
