Enhancing Joint Probabilistic Data Association with LSTM-Based Clutter Filtering for Improved Radar Target Tracking
Esra Alhadhrami, Clément Pira, Rami Kassab, Amal El Fallah Segrouchni, Frédéric Barbaresco · 2024
This paper presents an innovative approach to radar data processing by integrating Long Short-Term Memory (LSTM) networks with Joint Probabilistic Data Association (JPDA). Focusing on the challenge of reducing clutter in radar data, our method employs LSTM for its proficiency in sequential data analysis to filter out clutter and enhance measurements quality. Subsequently, JPDA is applied for precise target tracking. Through this synergy, we demonstrate the improvements in tracking accuracy, showcasing the potential of this approach in advancing radar technology. The paper provides a comprehensive evaluation of the method, including experimental results and comparison.