AI for Communication

NN-defined modulator

The project aims to revolutionize the field of wireless communications by reconstructing conventional wireless modulators with an interpretable neural network architecture. Such a new architecture will introduce a level of flexibility and portability that allows seamless operation across heterogeneous platforms. Also, by embedding AI-native learning capabilities directly on the device, this approach is set to redefine how modulation schemes are designed, implemented, and understood. 

Semantic communication

This project aims at developing new forms of communication beyond the conventional transceiver. It emphasizes the importance of meaning in the transmission and reception of messages other than the transmission and reception of signals. This involves not just the exchange of information, but the accurate conveyance and interpretation of the intent, context, and nuances behind the information. 

High-performance AI-native Distributed Unit on MEC 

This project aims to revolutionize the radio access network (RAN) ecosystem through the innovative integration of native AI design principles and edge computing technologies. By leveraging the capabilities of Multi-access Edge Computing (MEC), this project aims to significantly enhance the performance, efficiency, and intelligence of RAN operations and its associated applications, marking a substantial leap forward in telecommunications infrastructure.