Software Artificial Intelligence for Wireless Communications
In the future generation of Open Radio Access Networks (O-RANs), data are distributed among many wireless devices. Artificial Intelligence (AI) will encounter new challenges due to (i) the devices’ limited computational, bandwidth, and energy resources, (ii) the distributed data sets and software, and (iii) privacy and security. Current AI is borrowed from robotics, vision, or speech and text recognition, and is not conceived to meet such challenges.
SAICOM proposes a work plan around three main pillars: (1) new application-layer AI methods and algorithms for O-RAN; (2) new AI methods and algorithms for low-layer O-RAN protocols that heavily impact on the application layer; and (3) development of the complete AI software pipeline for AI over and within O-RAN.
