Machine Learning (ML), and specifically deep learning, has become a prominent and rapidly growing research topic within the field of wireless communications, both in academia and industry. In a discipline traditionally driven by compact analytic mathematical models, ML brings along a methodology that is data-driven and carries a major shift in the way wireless systems are designed and optimized. This brings with it both promise of more accurately representing complexities of the real world, as well as a great challenge in providing the same levels of analytic performance guarantee and validation we are used to in communications systems. While MLC has been already applied to self-organized networks, sensing, cognitive radio, resource allocation, and coding & modulation aspects, these have largely focused on more constrained tasks and learning environments. In more recent years, the algorithms, tools, computational power, availability of data, and other enablers have led machine learning to more directly solve for larger tasks and signal processing functions within communications systems. This mirrors the significant breakthroughs within ML in applications such as computer vision and natural language processing of embracing large datasets, concurrent tensor processing, and end-to-end (E2E) learning techniques providing solutions for high complexity tasks. An intriguing recent field is the design of highly flexible E2E solutions, where the whole communication model including the channel can be learned. Such designs combine superior ML-enabled transceiver design with data-driven channel and system identification. This may require distributed edge intelligence as well as privacy-preserving mechanisms. It is therefore natural to extend the investigations to the broader field of ML for communication, control, and security & privacy with its strong applications in a wide range of applications in 5G/6G, vehicular, AR/VR, IoT, and Tactile Internet among others. This workshop seeks to provide a first-tier platform for the dissemination of fundamental and applied research results as well as experimental demonstrations in such exciting fields of MLC.
Beyond providing a platform for the latest high-quality results in the field of machine learning for communication systems and encouraging fruitful and controversial discussions on the core challenges and prospect of the field, this workshop seeks to follow the main theme of the ICC 2020 version in promoting and encouraging open-ness, rigor, and reproducibility. As data measurement, processing, and learning systems are often significantly more intricate and specialized than compact analytic models, they often contain numerous details regarding the composition of the dataset, hyper-parameters and processing stages used within the learning and inference process, and countless additional implementation details which are difficult to compactly document within a concise and compact paper, but are easily captured within open software and data publications. This has become the norm in a number of machine learning-centric venues (e.g. NeurIPS, ICML), and rigorous new algorithmic work requires the publication and verification of open research. To embrace this within the IEEE ecosystem, this workshop is focused on directly supporting open-ness within machine learning for communications research, and asking researchers to share datasets, code, implementations, and baselines used throughout their work to help facilitate reproducibility and quantitative comparison by others within the field who may be able to critique, leverage, or extend research when it is conducted in such an open and reproducible manner.
As such, we invite the submission of novel, rigorous machine learning for communications research papers on new applications, ideas, and approaches along with the joint publication of datasets and source code required to reproduce the work by others. To mirror this spirit of openness and to help accelerate the research process, IEEE has offered to provide assistance hosting open datasets, software and pre-publication which allow for reproducing, modifying, and extending the work by many others. We invite authors to embrace widely used tools such as GitHub and/or GitLab for hosting their verifiable source code, baselines and implementations, embrace repositories such as ArXiv for early pre-publication feedback of works, and to embrace open-source tools such as GNU Radio and iPython notebooks.
Topics of Interest
We invite submissions of unpublished works on the application and theory of machine learning to communications. Below, we provide a non-exhaustive list of possible topics. We do not restrict the type of machine learning techniques.
- Machine learning driven design and optimization of modulation and coding schemes
- Machine learning techniques for channel estimation, channel modeling, and channel prediction.
- Machine learning based enhancements for difficult to model communications channels such as molecular, biological, multi-scale, and other non-traditional communications mediums
- Transceiver design and channel decoding using deep learning
- Machine learning driven techniques for radio environment awareness and decision making
- Machine learning for Internet of things (IoT) and massive connectivity.
- Machine learning for ultra-reliable and low latency communications (URLLC).
- Machine learning for Massive MIMO, active and passive Large Intelligent Surfaces (LIS).
- Machine learning for vision-aided wireless communications
- Distributed learning approaches for distributed communications problems
- (Deep) Reinforcement Learning and Policy learning for resource management & optimization
- Reinforcement Learning for self-organized networks and AP/BTS optimization
- Machine learning techniques for non-linear signal processing
- Low-complexity and approximate learning techniques and power reduction applications
- Machine learning for edge Intelligence, sensing platforms
- Algorithmic advances in machine learning for communication systems
- Advancing the joint understanding of information theory, capacity, complexity and machine learning communications systems
- Machine learning methods for exploiting complex spatial, traffic, channel, traffic, power and other distributions more effectively using measurement vs idealized distributions.
- Applications of transfer learning in wireless communication
- Compression of neural networks for low-complexity hardware implementation
- Unsupervised, semi-supervised and self-supervised learning approaches to communications
- Machine learning framework for joint communication, control, and security & privacy
- Privacy-preserving distributed machine learning for communications networks
Please submit your papers in accordance with IEEE GLOBECOM workshop paper standards via EDAS at https://edas.info/N27528
- Paper Submission Deadline: 14 August 2020
- Paper Acceptance Notification: 15 September 2020
- Camera-Ready: 1 October 2020
- Gerhard Wunder, Freie Universität Berlin, Germany (firstname.lastname@example.org)
- Elisabeth de Carvalho, Aalborg University, Denmark (email@example.com)
- Tim O’Shea, DeepSig, Arlington, VA, USA (firstname.lastname@example.org)
- Marios Kountouris, EURECOM, France (Marios.Kountouris@eurecom.fr)
- Zhi Ding, University of California, Davis, CA, USA (email@example.com)
- Jakob Hoydis, Nokia Bell Labs (firstname.lastname@example.org)
- Ahmed Alkhateeb, Arizona State University, USA (email@example.com)