A Comprehensive Investigation of Security Vulnerabilities in Distributed Multi-Robot Systems

The rapid deployment of multi-robot systems in critical missions such as exploration, surveillance, and disaster response has heightened the demand for cyber-connectivity, thereby increasing their exposure to cyber–physical threats. Among these applications, exploration serves as the foundation for subsequent tasks. Most existing exploration algorithms assume cooperative, trustworthy teammates and secure communication, overlooking insider threats and timing-based attacks. We investigate the vulnerability of distributed online exploration algorithms to two particularly stealthy attack types Sybil and Delay attacks, which exploit identity spoofing and message timing manipulation to degrade coordination, induce redundancy, and compromise coverage. Through simulation-based experiments on varying grid sizes and team configurations, we show that these attacks lead to incomplete mapping, false area coverage, and significant mission delays. Furthermore, strategic placement of adversarial robots is analyzed to evaluate the spatial influence of attacks. The study highlights critical vulnerabilities in current MRS coordination frameworks and motivates the design of self-adaptive, lightweight, and decentralized security mechanisms for resilient multi-robot exploration in adversarial environments.