A Review of Multi-Agent Coordination Strategies for Multi-Role UAV Swarms in Search and Rescue

Abstract

Unmanned aerial vehicle (UAV) swarms are increasingly investigated for search and rescue (SAR) operations to improve coverage, reduce response time, and enhance robustness in complex, time-critical environments. However, effective coordination remains a fundamental challenge due to communication constraints, dynamic task requirements, and the need to support heterogeneous roles. While many coordination strategies exist, a unified and SAR-focused comparative understanding of multi-agent coordination paradigms remains limited. This review analyzes 82 research studies, primarily published between 2015 and 2025, while incorporating selected foundational works published prior to 2015 to provide theoretical and historical context, focusing on coordination strategies for multi-role UAV swarms in SAR scenarios. A structured taxonomy is developed, covering swarm intelligence-based, consensus and graph-theoretic, market-based, learning-based, and hybrid approaches. These paradigms are systematically evaluated using a coordination-centric framework considering scalability, robustness, communication dependency, adaptability, and heterogeneity support. An illustrative SAR case study connects theoretical strategies with practical system behavior. The analysis shows that no single paradigm fully satisfies SAR requirements, highlighting trade-offs between coordination efficiency, communication overhead, and adaptability, alongside growing interest in hybrid approaches. This review provides structured insights to guide the design of scalable, robust, and adaptive multi-UAV SAR systems.