AI/ML-Assisted Positioning and Sensing in Wireless Network

The course will be developed in four modules.
Module 1: Introduction to AI/ML in Positioning

• Overview of traditional vs. ML-based positioning methods.
• Current standardization efforts for AI/ML-assisted positioning
• Introduction to Soft Information (SI) and its impact on positioning accuracy.
• Lab: Implement basic models for positioning.

Module 2: Soft Information and Feature Engineering

• Techniques for extracting and leveraging SI from signal and environmental data.
• Lab: Apply feature engineering and SI for improved localization accuracy and reduced communication and computation complexity.

Module 3: Anomaly Detection and Integrity Monitoring

• ML approaches for anomaly detection and integrity monitoring in positioning data.
• Lab: Develop models for anomaly detection to ensure reliable and robust position estimates.

Module 4: Reducing Latency and Improving Efficiency

• Use of ML to reduce the measurement overhead
• Advanced methods (e.g., belief condensation) to reduce latency and improve computational efficiency.
• Lab: Implement efficient ML models focused on real-time positioning with optimized complexity.

A minimum of 75% attendance is required.

Teaching methods

• Lectures with interactive discussions to cover theoretical concepts.
• Hands-on labs to practice implementing algorithms in MATLAB
• Case studies and group discussions on real-world applications, challenges, and industry standards (e.g., 3GPP).
• Q&A sessions to reinforce key topics and address student queries.