Localization and Tracking using ODE-LSTM Algorithm with Non-Line-of-Sight Channels
Xiao Hong Zhao, Feng Tian, Ziling Shao · 2023
In this study, we introduce a new method for localization and tracking in non-line-of-sight (NLOS) scenarios using a single Base Station (BS). By analyzing MIMO scatter channels, we create a geometric model that considers Angle of Arrival (AOA), Angle of Departure (AOD), and Time of Arrival (TOA) to determine positions. To enhance accuracy, we formulate an optimization problem accounting for time and directional biases in these relationships. For continuous tracking of mobile UEs, we propose an innovative ODE-LSTM algorithm that combines an Ordinary Differential Equation (ODE) solver with a Long Short-Term Memory (LSTM) network. This integration enables seamless tracking over various time intervals. Simulation results demonstrate the superiority of our approach in nonlinear tracking compared to traditional methods like Kalman filter(KF).