A Comprehensive Review of Deep Hyperbolic Graph Attention Network-Based Collaborative Routing Algorithm for Clustered IoT-MANETs

Abstract

The rapid advancement of Internet of Things (IoT) technologies has significantly increased the demand for efficient and intelligent routing mechanisms in Mobile Ad Hoc Networks (MANETs). IoT-enabled MANETs consist of dynamic, decentralized, and resource-constrained devices that communicate without fixed infrastructure, making routing a complex and challenging task. Traditional routing protocols such as AODV, DSR, and OLSR often struggle to maintain network efficiency due to frequent topology changes, node mobility, and limited energy resources. Recently, deep learning and graph-based networking approaches have emerged as promising solutions for improving routing performance in such environments. In particular, Graph Neural Networks (GNNs) and Graph Attention Networks (GATs) have demonstrated the ability to model complex network relationships and enable adaptive routing decisions. Furthermore, hyperbolic graph representation learning has gained attention for efficiently modeling hierarchical and scale-free network structures commonly observed in IoT-MANET environments. This review presents a comprehensive analysis of Deep Hyperbolic Graph Attention Network-based collaborative routing algorithms designed for clustered IoT-MANET systems. The study examines recent developments in clustering techniques, graph neural network architectures, hyperbolic embeddings, and intelligent routing strategies. It also compares existing approaches based on performance metrics such as network scalability, energy efficiency, packet delivery ratio, and routing overhead. The review highlights current research trends, identifies key challenges, and outlines future research directions for developing intelligent and scalable routing solutions for next-generation IoT-MANET networks.