Artificial Intelligence (AI) and Machine Learning (ML) approaches, well known from Computer Science disciplines, are beginning to emerge in the RF Signal Processing and Communications domains. However, there are various challenges arising in the application of Machine Learning to RF signals, such as inherently high data rates, sensitivity to environmental effects (noise, multi-path, interference etc), presence of multi-scale features in both frequency and time domains, to name a few. Also, in contrast to the image and text processing domains, the scarcity of large public repositories of standardized RF signal data makes it harder for academic and industry researchers to test and validate their algorithms in a robust, reproducible, and scalable fashion. Our goal is to organize a Conference workshop that would bring together researchers from the RF Signal Processing and Machine Learning communities, showcase state-of-the-art ML approaches applicable in the RF domain, and provide a forum for discussing cross-disciplinary ideas to address present and future challenges.

Some potential topics of interest include:

* ML for blind channel and signal characterization

* ML for source separation

* ML for RF signal classification

* ML for cognitive radio communications, e.g. optimization of spectrum usage dynamics
and spectrum access control

* Quality of unsupervised learning with corrupted, censored and missing spectrum
sensing samples

* Privacy-preserving ML for cognitive radio communications, e.g. in 5G cellular
networks

* ML for RF-based geolocation

* Distributed multi-agent learning in collaborative autonomous systems

* ML of the topology and structural properties of (cognitive) radio networks

* Reinforcement learning in wireless networks

For more information visit https://ijcai19.org/