Home // ADAPTIVE 2024, The Sixteenth International Conference on Adaptive and Self-Adaptive Systems and Applications // View article

Agent-based Modeling in the Edge Continuum using Swarm Intelligence

Authors:
Melanie Schranz
Kseniia Harshina
Peter Forgacs
Fred Buining

Keywords: Swarm Intelligence, Bio-inspired Algorithm, Edge Continuum, Agent-Based Modeling

Abstract:
The edge continuum presents a dynamic and evolving paradigm in the future's world of computing, offering a versatile and efficient solution for a wide range of applications and industries. The edge infrastructure is more challenged in its stability and performance because of more stringent latency and autonomy requirements, distribution across multiple sites, their local limited size, multi-tenancy and multi-operators, local management, with components being concurrent and asynchronous. This paper introduces an innovative framework that combines agent-based modeling and swarm intelligence to address complex challenges such as resource allocation, workload scheduling, and data management in the edge continuum. This framework, at the core of the architecture, enhances edge autonomy, reduces latency, improves energy efficiency, and optimizes cloud connectivity by applying agent-based modeling. By integrating autopoietic characteristics like self-organization, regeneration, and regulation, the system dynamically adapts to changing conditions. Two candidate algorithms, the hormone algorithm and ant algorithm, emulate decentralized decision-making processes observed in nature. The paper reviews related work in swarm intelligence for network optimization and emphasizes the need for distributed, agent-based solutions. This research paves the way for robust, adaptive, and scalable systems in the complex edge environment, promising emergent behaviors and enhanced efficiency. In this position paper, we propose the edge continuum with its characteristics and limitations as a novel field of application for swarm intelligence by conceptually proposing agent-based modeling and simulation.

Pages: 1 to 7

Copyright: Copyright (c) IARIA, 2024

Publication date: April 14, 2024

Published in: conference

ISSN: 2308-4146

ISBN: 978-1-68558-153-4

Location: Venice, Italy

Dates: from April 14, 2024 to April 18, 2024