In recent years, there has been a trend of companies sending more and more satellites into orbit. More satellites in orbit mean more ground station installations on earth, which inherently implies more space data traffic.
The inflation in satellite communication significantly increases the chances of misconfigurations and misalignments at the ground terminal operator trying to coordinate all this communication. As a result, radio frequency interference became a severe revenue-threatening problem for satellite operators.
Edgise partnered with Antwerp Space and The European Space Agency (ESA) to design an intelligent AI-based satellite signal processing system capable of detecting the presence of interference, classifying the type of interference, and predicting the best modulation coding scheme for the communication link. Edgise provides edge artificial intelligence knowledge for designing the AI algorithms and shares expertise for several off-the-shelf AI chipsets and available edge AI hardware systems.
"Signal quality is crucial in space. This onboard AI system will empower satellites to predict the quality better and allow operators to solve the interference problem quicker," according to Nick Destrycker (founder of Edgise).
Edgise will design the AI system so it can retrain itself autonomously onboard the satellites, removing the need for the operator to constantly send AI model updates to the satellite, wasting precious up and downlink bandwidth.
After the current project, Edgise and Antwerp Space plan a follow-up trajectory to test the system in space during an in-orbit demonstration.