Abstract
<jats:p>With the continuous development of wireless communication technologies, the integration of Ultra-Wideband (UWB) and Inertial Navigation Systems (INS) has become a widely adopted solution for indoor navigation. UWB provides high positioning accuracy, while INS ensures continuous estimation, and their fusion effectively compensates for the limitations of each system. Nevertheless, in collaborative navigation involving multiple unmanned vehicles, challenges such as communication delays, packet loss, and imperfect mechanisms for weight allocation and anomaly detection in information fusion still restrict overall accuracy and stability. To address these issues, this paper proposes a packet-delay-optimized UWB–INS integrated navigation method. Specifically, a buffering mechanism is introduced to cache and schedule asynchronous data, ensuring temporal consistency in the fusion process, while an entropy-based weighting method dynamically adjusts the contribution of each information source according to its uncertainty. Experimental results conducted in complex indoor environments demonstrate that the proposed method can effectively reduce positioning errors caused by packet delay and loss, significantly improve accuracy and robustness, and provide reliable technical support for practical applications of unmanned vehicles in challenging environments.</jats:p>