Skip to main content


Deep learning (DL) has appeared as a promising tool to re-design the physical layer by adding intelligence to the wireless networks requiring new features for future communications, such as large-scale topology, accurate processing, and high-speed requirements. These complex features present an opportunity for the success of DL-based frameworks because of the lack of mathematical models, the conventional methods fall short in practice. DL empirically fits the neural network model to the raw data and recent results have been extremely promising.
In this talk, we present the key issues of the wireless communication system design. Then, we show the potential of DL frameworks that aims to design all-in-one solutions. Later, we present challenges in designing DL-based physical layer communication, such as latency and space constraints, black-box nature, re-training issues, and data collection challenges. We also present recent applications of DL-based intelligent wireless networks, such as redesigning the conventional signal processing blocks for modulation, channel coding, and detection by replacing the communication system with an autoencoder for both the wireless and optical networks and performing mm-wave path-loss prediction for complex urban environments. We also focus on showing how to open the black-box of the DL-based designed wireless networks. Lastly, we present exciting new learning solutions for designing physical layer communications, such as deep reinforcement learning and model-based learning.

Mathini SellathuraiBio

Mathini Sellathurai is currently a professor of signal processing and wireless communications and Dean of Science and Engineering with Heriot-Watt University, Edinburgh, U.K. She has been active in signal processing research for the past 20 years and has a strong international track record in wireless communications. She held visiting positions with Bell-Laboratories, Holmdel, NJ, USA, and at the Canadian Communications Research Centre, Ottawa, Canada. She has published over 200 peer reviewed papers in leading international journals and conferences, given invited talks and has written several book chapters as well as two research monographs. Her present research includes machine learning and statistical signal processing techniques applied to wireless communications, radar, and medical applications. She is a recipient of an IEEE Communication Society Fred W. Ellersick Best Paper Award (2005), the Industry Canada Public Service Awards for contributions to Science and Technology (2005), and Awards for contributions to Technology Transfers to Industry (2004). She was the recipient of the Natural Sciences and Engineering Research Council of Canada (NSERC) Doctoral Award for her Ph.D. dissertation in 2002. She was an Editor for IEEE TRANSACTIONS ON SIGNAL PROCESSING from 2009 to 2014, and from 2015 to 2018, the General Chair of IEEE Signal Processing Advances in Wireless Communications (2016) in Edinburgh, and a member for IEEE SPCOM Technical Strategy Committee from 2014 to 2019. Presently she is a member of Strategic Advisory Board of Engineering and Physical Sciences Research Council, UK, for Information and Communications Technologies.