New Radio User Equipment for Positioning Accuracy Enhancement using AI

Positioning accuracy is crucial for New Radio (NR) User Equipment (UE) in various applications, including navigation, emergency response, and IoT-enabled services. However, environmental factors such as multipath propagation and non-line-of-sight (NLOS) conditions in urban and indoor environments often degrade the positioning accuracy of NR systems. To address these challenges, this paper presents an AI-based approach for enhancing NR UE positioning accuracy. The proposed method leverages a neural network model trained on noisy NR positioning data to predict corrected positions, mitigating the impact of common errors in real-world scenarios like complex and dynamic environments. By implementing machine learning models for signal fingerprinting, sensor fusion, beam fusion, beam prediction and mobility estimation, but in proposed system intelligently adapts to varying channel conditions and user contexts. Simulated data with line-of-sight (LOS) and NLOS conditions are used to test the model's performance, demonstrating a significant improvement in positioning accuracy compared to conventional methods. Results indicate that the AI-driven model significantly mitigates positioning error, thereby enhancing the reliability and precision of NR UE positioning. This innovative solution harbors promising applications in 5G and beyond, contributing to the evolution of accurate, AI-enabled location-based services.