TU Berlin

Communications and Information Theory ChairInvitation for a talk by Prof. Upamanyu Madhow, Ph.D., University of California, Santa Barbara, U.S.

Schriftzug CommIT

Page Content

to Navigation

Invitation for a talk by Prof. Upamanyu Madhow, Ph.D., University of California, Santa Barbara, U.S.

TITLE:
Millimeter Wave Systems: Where Signal Processing meets Hardware & Physics
TIME:
Sept 03, 2019, 11:00 a.m.
PLACE:
Room HFT-TA 617, Einsteinufer 25, 10587 Berlin

ABSTRACT: Utilizing the essentially unlimited bandwidth and spatial reuse at millimeter wave and THz frequencies requires the development of signal processing approaches that exploit the physics of tiny wavelengths, while accounting for hardware constraints associated with scaling bandwidth and spatial resolution. In this talk, we start with an overview of the evolution of research at UCSB in the area of millimeter wave communication and sensing, and our current efforts within a large multi-university center led by UCSB. We then provide a more detailed account of our recent work on all-digital mmWave massive MIMO.  Key observations are as follows: (1) using tiled architectures, one RF chain per antenna element is indeed possible even as we scale up the number of antenna elements, making it possible to scale up the number of simultaneous users proportionally with all-digital MIMO signal processing; (2) when we take advantage of the scale enabled by a large number of antennas, a number of hardware constraints can be relaxed, including linearity in the RF chain, phase noise specifications, and the precision of analog-to-digital conversion; (3) for the sparse channels characteristic of these bands, it is possibly to greatly simplify the MIMO signal processing required for supporting multiple simultaneous users.

BIO: Upamanyu Madhow is Distinguished Professor of Electrical & Computer Engineering at the University of California, Santa Barbara. His current research interests focus on next generation communication, sensing and inference infrastructures centered around millimeter wave systems, and on robust machine learning. Dr. Madhow is a recipient of the 1996 NSF CAREER award, and co-recipient of the 2012 IEEE Marconi prize paper award in wireless communications. He is the author of two textbooks published by Cambridge University Press, Fundamentals of Digital Communication (2008) and Introduction to Communication Systems (2014).  He has been heavily engaged in technology transfer of his work through industry collaborations, and as co-founder of several startups.

Invitation for a talk by Dipl.-Ing. Karl-Ludwig Besser, Technische Universität Dresden, Germany

TITLE:
Wiretap Code Design by Neural Network Autoencoders
TIME:
Sept 03, 2019, 02:00 p.m.
PLACE:
Room HFT-TA 617, Einsteinufer 25, 10587 Berlin

ABSTRACT: In industrial machine type communications, an increasing number of wireless devices communicate under reliability, latency, and confidentiality constraints, simultaneously. From information theory, it is known that wiretap codes can asymptotically achieve reliability (vanishing block error rate (BLER) at the legitimate receiver Bob) while also achieving secrecy (vanishing information leakage (IL) to an eavesdropper Eve). However, under finite block length, there exists a tradeoff between the BLER at Bob and the IL at Eve. In this presentation, it is shown how neural network autoencoders can be used to flexibly design finite blocklength wiretap codes. To attain this goal, a multi-objective programming problem is formulated, which takes the BLER at Bob and the IL at Eve into account. Simulation results show that the proposed scheme can find codes outperforming polar wiretap codes with respect to both BLER and IL simultaneously.

BIO: Karl-Ludwig Besser received his Dipl.-Ing. degree in electrical engineering from Technische Universität Dresden in 2018. In August 2018, he joined the Communications Theory group at TU Dresden. His research interests are in the area of physical layer security and the application of machine learning in communications.

Navigation

Quick Access

Schnellnavigation zur Seite über Nummerneingabe