ELIAM: An Ellipsoid Windowing-Based Dataset Generation Framework for LiDAR and Position Aided mmWave Communications

Sensing-aided communication is a key element of future 6G networks, with LiDAR widely used in mmWave vehicular communications for its precise geometric data. However, the limited field of view of vehicle-mounted LiDAR often results in sparse roadside point clouds. To address this, we propose ELIAM, an Ellipsoid Windowing-based dataset generation framework. ELIAM integrates aerial data collection to enrich sparse, real-time vehicle LiDAR data with dense point clouds of stationary roadside objects. The Ellipsoid Windowing technique further optimizes the dataset by filtering irrelevant points, reducing size while retaining critical information. Experimental results confirm that ELIAM enhances dataset completeness and improves beam alignment accuracy for mmWave communications.