Adaptive Step Length Estimation and GNSS Integration to Improve PDR Accuracy

This paper proposes a novel method to improve the accuracy of Pedestrian Dead Reckoning (PDR) by refining step length estimation and integrating GNSS data using Kalman filters. We developed an adaptive step length estimation model that dynamically adjusts based on real-time gait state detection. Our experimental results demonstrate that the proposed method significantly enhances PDR accuracy across various gait patterns and environments, providing a more robust and reliable positioning solution for smartphone users.