Khalifa University of Science and Technology, the GSM Association (GSMA) and AMD are launching an operator evaluation initiative for RFGPT during MWC Doha 2026, giving telecommunications operators an opportunity to assess the artificial intelligence model’s capabilities and applications in radio frequency analysis.
The collaboration combines Khalifa University’s research expertise in AI and wireless communications, GSMA’s global mobile industry ecosystem and AMD’s advanced computing capabilities to explore how foundation models can address emerging challenges in telecommunications networks.
Developed by Khalifa University’s Digital Future Institute, RFGPT is a Radio Frequency (RF) foundation model that combines wireless signal analysis with natural-language interaction. It allows users to ask questions about observed wireless signals, identify their characteristics and interpret complex environments containing multiple signals.
The model is designed to make technical radio frequency data easier and faster to interpret, with potential applications in radio access network (RAN) optimisation, network planning, spectrum management and interference analysis. It will be hosted on AMD computing infrastructure and made available to operators through GSMA-supported access.
RFGPT supports six evaluation areas. These include identifying wireless technologies such as Wi-Fi, Bluetooth, LTE, 5G New Radio (NR), UMTS and DVB-S2, as well as distinguishing between uplink and downlink transmissions where applicable.
The model can also recognise modulation schemes, analyse wideband environments containing multiple signals and assess signal coexistence by identifying time-frequency overlap between transmissions. It can estimate signal interaction and interference levels to support network analysis.
Additional capabilities include assessing WLAN and NR user densities within observation windows to help operators understand network utilisation patterns. RFGPT can also interpret key 5G NR parameters, including subcarrier spacing, synchronisation signal block (SSB) patterns, channel state information reference signal (CSI-RS) configurations and sounding reference signal (SRS) structures.
Khalifa University Senior Director of the Digital Future Institute, Professor Merouane Debbah, said the initiative would give operators an opportunity to explore radio frequency foundation models in practical settings and provide feedback to guide their development.
GSMA Director of AI Technologies Louis Powell said the evaluation would help determine where the model could deliver value and what would be required to advance its applications.
AMD Executive Vice President and Chief Commercial Officer Philip Guido said the collaboration would allow operators to assess how AI could simplify radio frequency analysis and support network planning, spectrum management and interference mitigation.
Telecommunications operators interested in participating are invited to submit their use cases. The research and hosting teams will coordinate evaluation scope and access arrangements, with participants expected to provide feedback on the model’s performance, usability and relevance to their applications.
