AI System Shows Tenfold Gain in Blocking Satellite Interference
Results Are Preliminary, Company Says, With Further Testing Still Ahead
An artificial intelligence system detected and canceled satellite signal interference up to 10 times more effectively than conventional methods in early testing, according to results satellite communications provider Gilat Satellite Networks announced Thursday.
“In an environment where spectrum is becoming a crucial resource for both defense and commercial applications, this technology opens new opportunities that were not possible before.”
Aharon Mullokandov, Gilat
The Petach Tikva, Israel-based company, which trades on the Nasdaq and the Tel Aviv Stock Exchange as GILT, said the patent-pending system runs AI-based signal processing on standard edge neural processing hardware. Gilat said it tested the system against interference and jamming patterns that were neither fixed nor cooperative — the kind of unpredictable signals that are harder for older mitigation tools to track.
The results are preliminary and come from Gilat’s own testing rather than an independent evaluation. The company said the target signal itself showed only negligible quality loss after the interference was canceled, but that the system still needs to be demonstrated on dedicated AI accelerator hardware before it could be built into operational satellite terminals.
“In an environment where spectrum is becoming a crucial resource for both defense and commercial applications, this technology opens new opportunities that were not possible before,” said Aharon Mullokandov, Gilat’s chief R&D officer.
Gilat said it plans to move the system onto dedicated AI accelerator hardware next, which the company said would allow a lower-power version suitable for terminals already fielded in commercial and defense satellite networks.
The company did not give a timetable for when the technology might reach commercial satellite terminals.