Historical CSI
10 channel instants before the prediction origin.
Ground-level XL-MIMO Dataset
This page visualizes a representative part of our current Sionna RT ground-level dataset. It shows an edge-transition trajectory, synchronized sensing views, and measured CSI evolution across three propagation scenarios.
7Scene layouts
64 × 64Physical array
10 + 12Historical + future CSI
10 msSensing interval
Scene and Mobility Visualization
A saved view of the ground-level XL-MIMO dataset shows the layout, user movement, Sionna RT propagation, and synchronized RGB and LiDAR sensing for one representative trajectory.
Multimodal Trajectory Visualization
Move through the saved frames of trajectory L00_edge_v20_000. Each frame shows the ray-tracing and sensing views captured at the selected instant.
Edge transition t = 0 ms
0 ms
Future sensing frames are displayed for dataset inspection only. They are not available to a one-shot predictor at t = 0.
Wireless Channel Visualization
These are CSI-to-CSI properties of the dataset across prediction horizons, not prediction NMSE. The curves summarize 797 trajectories per scenario.
Channel correlation
Median normalized complex-channel correlation, relative to the prediction start.
Stable LoSStable NLoSEdge transition
Statistics use stored CSI at 1, 5, 10, ..., 100 ms. This diagnostic cohort does not cover every trajectory in the complete dataset.
Dataset Organization
The complete CSI and sensor files remain in dataset storage; this page provides a compact, verified public preview.
Future CSI
12 target horizons from 1 to 100 ms.
Environment Sensing
RGB, LiDAR and position records with timestamps.
Propagation Metadata
Scenario, layout, speed and path visibility where available.
Data sources: channel-evolution summaries, LoS-visibility records, and saved ground-level XL-MIMO dataset captures. No curve on this page represents a model result.