 Co-operative localization, CL, has become increasingly important in the field of robotics and unmanned aerial vehicles, UAVs. It is now being applied to groups of firefighters who are able to work together in formation. Previous studies have focused on pedestrian localization, which is not possible with UAVs or robots due to their ability to form a team. This study developed an adaptive decentralized cooperative localization, ADCL, algorithm for a group of firefighters. Each member maintained a local filter and the relative measurement noise covariance was estimated instead of a fixed value. Expressions were derived for the intermember collaboration, reducing the influence of NLOS errors in ultra-wideband, UWB, ranging. Two experiments were conducted in a building and forest environment, demonstrating improved accuracy and suppression of localization errors by 14.23% and 47.01%, respectively, when compared to the Dklechf algorithm. This article was authored by Yang Chong, Xiang Guosu, Ning Yinghua, and others. We are article.tv, links in the description below.