Agent-Based Modelling Approach for Distributed Decision Support in an IoT Network

TL;DR

This paper explores the use of Agent-Based Modeling to simulate and analyze communication protocols in IoT networks, specifically in traffic management, highlighting the benefits of coordination among decision makers.

Contribution

It demonstrates the application of ABM to model IoT communication protocols and compares different MAC layer strategies, including centralized, desynchronization, and decentralized learning approaches.

Findings

Coordination improves information accuracy and spectrum utilization.

ABM effectively models heterogeneous IoT network interactions.

Different MAC protocols have varying impacts on network performance.

Abstract

An increasing number of emerging applications, e.g., internet of things, vehicular communications, augmented reality, and the growing complexity due to the interoperability requirements of these systems, lead to the need to change the tools used for the modeling and analysis of those networks. Agent-Based Modeling (ABM) as a bottom-up modeling approach considers a network of autonomous agents interacting with each other, and therefore represents an ideal framework to comprehend the interactions of heterogeneous nodes in a complex environment. Here, we investigate the suitability of ABM to model the communication aspects of a road traffic management system, as an example of an Internet of Things (IoT) network. We model, analyze and compare various Medium Access Control (MAC) layer protocols for two different scenarios, namely uncoordinated and coordinated. Besides, we model the scheduling mechanisms for the coordinated scenario as a high level MAC protocol by using three different approaches: Centralized Decision Maker, DESYNC and decentralized learning MAC (L-MAC). The results clearly show the importance of coordination between multiple decision makers in order to improve the accuracy of information and spectrum utilization of the system.