MAPGRID_AI

What Is MAPGRID_AI?

What is it?

MapGrid_AI is a way to figure out where wireless transmistters and receivers are based on the signal strength and clarity (RSSI and SNR) without knowing any GPS coordinates.

MapGrid_AI can be used to map networks where actors have incentives to disguise their position. With enough datapoints, MGAI builds a relational or "latent" map. Think of it like a rigid piece of cloth shaped to terrain and structures in such a specific way that it maps to just one place on earth.

MGAI uses a custom built model trained on hundreds of cities and millions of RF readings, updating itself based on new information coming in.

MGAI is designed to interact with and learn from the real physical world, constantly improving based on results.

How does it work?

The MGAI demo works by ingesting multiple packets from transmitters traveling around an area. Using the RSSI and SNR of the packets in relation to other readings from other packets, MGAI builds a latent map of the relational positions of gateways.

At a threshold of datapoints, MGAI builds not only a map of gateaway positions, but also identifies physical and network topology into a unique fingerprint of the area.