Multiperson Cooperative Navigation Based on Geometric Constraints of Foot-to-Foot and Person-to-Person Distances

Maximum two-feet distance constraint and ultrawideband (UWB) technology are commonly used in pedestrian inertial navigation, but they suffer from poor precision and preinstalled anchors. In this work, a multiperson cooperative navigation method based on geometric constraints of foot-to-foot and person-to-person distances measured by UWB is proposed. Combined with gait analysis in kinematics, the gait is divided into the stance and swing phases. For each phase, attitude, position, and the related characteristics of two feet are analyzed and modeled. Furthermore, considering the influence of non-line-of-sight (NLOS) conditions on UWB ranging, the person-to-person distances are formed into the reticulate geometric constraint. An extended Kalman filter is established to estimate the attitude and position of each agent optimally according to the geometric constraints and position vectors. Comparative experiments are designed in various environments. Compared with the traditional inertial navigation system (INS) method, the proposed method improves the end-to-end error by 40%–70% and the average error by 35%–72%. The results show that the proposed method has higher accuracy and better robustness to the environments than the traditional INS method.