Wireless Future for 2020 and Beyond


Information Processing and Learning over Networks

Organizers and Chairs:

Kostas Berberidis, University of Patras, Greece
Dimitris Berberidis, Carnegie Mellon University, USA

Scope of the Papers

Modern networks consist of an exponentially increasing number of heterogeneous devices with numerous complex interactions taking place between them either explicitly or implicitly. Mining information and looking for structure patterns in the huge volumes of generated data is of paramount importance as it is expected to lead to various societal improvements. However, a critical re-examination of the existing information processing tools must be performed, since the individual devices are not able to handle the immense data, nor are they in a position to reveal and exploit the underlaying dependencies at different scales of the network.

To this purpose, the Internet of Things (IoT), which has been established as a powerful technology allowing different devices to take active part in the network, needs to be further enhanced with distributed and/or decentralized processing and learning capabilities. Devices with limited computing capabilities and communication resources will be able to collaboratively process information (and, for example, learn a model) without the need to send all data to the Cloud and then wait for the learning outcome to be fed back. In this spirit, new paradigms have recently been proposed, such as Edge Computing or Fog Computing, and D2D communications.

This Special Session aims to contribute to the ongoing research efforts for developing new tools and algorithms that will allow heterogeneous devices with possibly different tasks and interests to collaborate by various means, thus offering enhanced learning performance, affordable complexity, latency and robustness against attacks, without sacrificing security and privacy.

Topics of interest include (but are not limited to):

  • Distributed algorithms for signal processing tasks (e.g., channel estimation, beamforming, localization, signal enhancement, etc.)
  • Collaborative information processing and learning over graphs
  • Game theoretic approaches for coalition formation
  • Adaptive and/or distributed caching
  • Privacy preserving distributed learning
  • Edge intelligence and distributed IoT systems
  • Distributed machine learning at wireless network edge
  • Distributed synchronization and network organization
  • Federated learning with communication constraints

Submission Guidelines

The Special Session will include not only invited papers but also relevant papers from the Open Call. Full papers and short papers/extended abstracts as defined below can be submitted.

Full Papers: Full paper submissions of original work (not previously published, or under review at another conference or journal) must not be longer than five pages and will be published in the conference proceedings.

Short Papers and Extended Abstracts: : Submissions must not be longer than two pages. They should convince the reader that the author(s) would give an exciting presentation and stimulate lively discussion (will be published in the conference proceedings). Note that it is fully expected that extended abstract papers accepted for the session will eventually be extended as full papers suitable for formal academic publication and presentation at other conferences/publications.

Please use the IEEE template as described here. Accepted papers will be published in the conference proceedings and submitted to IEEE Xplore (approval pending).

Submissions are now accepted through EDAS: [Start a new submission here]

Important Dates

Paper submission deadline:   TBD
Notification of acceptance:     TBD
Camera-ready papers due:    TBD


Contact Us

Kostas Berberidis, University of Patras, Greece, Email: berberid@ceid.upatras.gr
Dimitris Berberidis, Carnegie Mellon University, USA, Email: berberidisd@gmail.com