Prof. Elza Erkip of NYU Tandon School of Engineering

Traditionally, the goal of compression is to represent a complex information source such as an image in the most compact way while ensuring an acceptable level of signal distortion. The goal of communication, on the other hand, is to reliably transmit compressed information over a noisy channel. This talk explores how in-network compression can be used for faster and more reliable communication in 6G networks. This is achieved by task-oriented compression, where instead of minimizing the signal distortion, the goal is to optimize a task described by the wireless network operation. By leveraging recent advances in learning-based data compression, we illustrate the potential benefits of learned task-oriented compression for 6G. We focus on a precoding-oriented channel state information (CSI) feedback scheme for multi-cell multi-user MIMO systems, where the learned end-to-end architecture integrates the downlink channel estimation, the CSI compression, and the downlink precoder for higher rates and more effective inference management. 

Biography:

Prof. Elza Erkip, IEEE Fellow is an Institute Professor in the Electrical and Computer Engineering Department at New York University Tandon School of Engineering. She received the B.S. degree in Electrical and Electronics Engineering from Middle East Technical University, Ankara, Turkey, and the M.S. and Ph.D. degrees in Electrical Engineering from Stanford University, Stanford, CA, USA. Her research interests are in information theory, communication theory, and wireless communications.

Dr. Erkip is a member of the Science Academy of Turkey and is a Fellow of the IEEE. She received the NSF CAREER award in 2001, the IEEE Communications Society WICE Outstanding Achievement Award in 2016, the IEEE Communications Society Communication Theory Technical Committee (CTTC) Technical Achievement Award in 2018, and the IEEE Communications Society Edwin Howard Armstrong Achievement Award in 2021. She was the Padovani Lecturer of the IEEE Information Theory Society in 2022. Her paper awards include the IEEE Communications Society Stephen O. Rice Paper Prize in 2004, the IEEE Communications Society Award for Advances in Communication in 2013 and the IEEE Communications Society Best Tutorial Paper Award in 2019. She was a member of the Board of Governors of the IEEE Information Theory Society 2012-2020, where she was the President in 2018. She was a Distinguished Lecturer of the IEEE Information Theory Society from 2013 to 2014. She is currently the Editor-in-Chief of IEEE Journal on Selected Areas in Information Theory and the Chair of IEEE Communications Society Communication Theory Technical Committee.

Decentralized Federated Learning with Mode

Email:

Yong Liu of NYU Tandon School of Engineering

Federated Learning (FL) trains a shared model using data and computation power on distributed agents coordinated by a central server. Decentralized FL (DFL) utilizes local model exchange and aggregation between agents to reduce the communication and computation overheads on the central server. However, when agents are mobile, the communication opportunity between agents can be sporadic, largely hindering the convergence and accuracy of DFL. In this talk, we present Cached Decentralized Federated Learning (Cached-DFL) to investigate delay-tolerant model spreading and aggregation enabled by model caching on mobile agents. Each agent stores not only its own model but also models of agents encountered in the recent past. When two agents meet, they exchange their own models as well as the cached models. Local model aggregation utilizes all models stored in the cache. We theoretically analyze the convergence of Cached-DFL, explicitly considering the model staleness introduced by caching. We design and compare different model caching algorithms for different DFL and mobility scenarios. We conduct detailed case studies in a vehicular network to systematically investigate the interplay between agent mobility, cache staleness, and model convergence. In our experiments, Cached-DFL converges quickly, and significantly outperforms DFL without caching.

Biography:

Prof. Yong Liu, IEEE Fellow is currently a Professor of the ECE department at the Tandon School of Engineering, New York University, USA. He is also affiliated with CATT and NYU WIRELESS. He obtained his Ph.D degree from the ECE Dept. at the University of Massachusetts, Amherst in 2002. His general research interests lie in modeling, design and analysis of networked systems. His current research directions include multimedia networking, resilient edge networks, decentralized learning over networks, and network measurement and analytics. He is the winner of outstanding junior alumni award of College of Engineering, University of Massachusetts, Amherst, 2017, the Best Paper Award at ACM Internet Measurement Conference (IMC) 2012, the National Science Foundation Career Award in 2010, the Best Paper Award of IEEE Conference on Computer Communications (INFOCOM) in 2009, and the IEEE Communication Society Multimedia Communications Best Paper Award in 2008. He is a fellow of IEEE for his contribution to multimedia networking since 2017.

Email: