T11: Why OTFS 2.0? – How to Enable AI/ML in 6G, How to Integrate Sensing and Communications
Co-chair: Saif Khan Mohammed, Indian Institute of Technology Delhi, India
Co-chair: Ronny Hadani, University of Texas at Austin, USA
Co-chair: Ananthanarayanan Chockalingam, Indian Institute of Science Bangalore, India
Co-chair:Robert Calderbank, Duke University, USA
Abstract: Machine Learning (ML) methods have revolutionized image and natural language processing and can be expected to improve traditional methods of communication based on mathematical models. However, 6G presents a deeper opportunity to reflect on the fundamentals of wireless communication, as it becomes more and more difficult to estimate, track and equalize channels with Doppler spreads measured in KHz. This tutorial is not organized around the question of how ML can improve traditional methods of wireless communication. It addresses the more fundamental question of what are the characteristics of a physical layer that ML algorithms want to see.
The tutorial introduces Orthogonal Time Frequency Space (OTFS) 2.0 modulation which is designed to make communication and radar sensing more predictable when compared to OTFS 1.0. OTFS 1.0 (or MC-OTFS) was designed as an OFDM overlay, and has been the focus of most research attention so far. We emphasize predictability since this is the characteristic that improves performance of ML algorithms and is fundamental to the question of whether functions like scheduling can move from the base station to the cloud. The OTFS 2.0 carrier waveform is a pulse in the delay-Doppler (DD) domain, formally a quasi-periodic localized function with specific periods along delay and Doppler. Viewed in the time domain, it is realized as a pulse train modulated by a tone (pulsone). We begin by showing that when the channel delay spread is less than the delay period, and the channel Doppler spread is less than the Doppler period, the OTFS 2.0 input-output (I/O) relation is predictable and non-fading. Given the I/O response at one DD point in a frame, it is possible to predict the I/O response at all other points. It is possible to learn the I/O relation without estimating the channel, opening up the possibility of a model-free mode of operation. The counterpart of predictability in radar sensing is unambiguous delay-Doppler estimation.
The tutorial then describes how filters in the discrete delay-Doppler (DD) domain can be used to design noise-like waveforms that enable joint sensing and communication. It then describes how to design transmit filters that optimize the tradeoff between predictability and time/bandwidth expansion. Finally, it considers tracking of multiple targets in radar applications to demonstrate the advantages of the OTFS 2.0 waveforms.
The tutorial aligns with growing international interest in spectrum sharing and new waveforms for 6G.
Program:
The tutorial is organized into six sections. The content of each section and the duration are listed
below.
• Section 1: Introduction (15 mins)
– OTFS 2.0: History and Future Perspective
– OTFS 2.0 vs OFDM – Preventing ISI vs. Acquiring ISI
– OTFS 2.0 to Enable AI/ML and Integrated Sensing and Communications in 6G
• Section 2: OTFS 2.0 – How to Make Communication and Radar More Predictable in 6G (45 mins)
– Representing Doubly Spread Channels in the Delay-Doppler Domain
– OTFS 2.0 Modulation – TD Representation of Pulses in the DD Domain
– Introducing the Radar Ambiguity Function – Quantifying Resolvability
– Predictability of the I/O Relation in OTFS 2.0 by Avoiding Aliasing in the DD Domain
– OTFS 2.0 for Model-Free Operation – Enabling AI/ML in 6G
• Break: 15 mins
• Section 3: OTFS 2.0 for Joint Sensing and Communications (45 mins)
– Separating Sensing and Communications – Baseline of Separate Zak-OTFS Subframes
– Integrating Sensing and Communications with Point Pulsones – Dividing Resources
– Filters in the Discrete Delay-Doppler Domain – Designing Noise-Like Waveforms
– Integrating Sensing and Communications with Spread Pulsones – Coexistence
• Section 4: Filter Design in OTFS 2.0 (20 mins)
– Designing Factorizable Filters in the Discrete Delay-Doppler Domain
– Optimizing the Tradeoff between Predictability and Time-Bandwidth Expansion
• Section 5: OTFS 2.0 for Radar (20 mins)
• Section 6: Conclusions (5 mins)
Co-chair Bios:
Saif Khan Mohammed
Bio: Saif Khan Mohammed is a Professor with the Department of Electrical Engineering, Indian Institute of Technology Delhi (IIT Delhi). He currently holds the Jai Gupta Chair at IIT Delhi. He received the B.Tech. degree in Computer Science and Engineering from IIT Delhi, New Delhi, India, in 1998, and the Ph.D. degree from the Electrical Communication Engineering Department, Indian Institute of Science, Bangalore, India, in 2010. From 2010 to 2011, he was a Post-Doctoral Researcher at the Communication Systems Division (Commsys), Electrical Engineering Department (ISY), Linkoping University, Sweden. He was an Assistant Professor at Commsys, from September 2011 to February 2013. His main research interests include, waveforms for high mobility scenarios in sixth generation (6G) communication systems, wireless communication using large antenna arrays, coding and signal processing for wireless communication systems, information theory, and statistical signal processing. He currently serves as an Editor for the IEEE Transactions on Wireless Communications and in the past he has served as an Editor for the IEEE Wireless Communications Letters and Physical Communication journal (Elsevier). He holds four granted U.S. patents and three granted Indian patents in his area of research. He received the 2017 NASI Scopus Young Scientist Award and the Teaching Excellence Award at IIT Delhi for the year 2016–2017. He was also a recipient of the Visvesvaraya Young Faculty Fellowship from the Ministry of Electronics and IT, Government of India, from 2016 to 2019.
Ronny Hadani
Bio: Ronny Hadani is an associate professor in the Mathematics Department of the University of Texas at Austin. Prior to that he was a Dickson post-doctoral fellow in the University of Chicago. His field of expertise is representation theory and harmonic analysis. Hadani is a co- founder of Cohere technologies and currently serves as its Chief Science Officer. Hadani is a coinventor of the OTFS modulation technique and has been granted over 70 OTFS related patents. Hadani holds a PhD in pure mathematics from Tel-Aviv University under advisory of Professor Joseph Bernstein and a Master degree in applied mathematics from The Weizmann Institute of Science under advisory of Professor David Harel.
Ananthanarayanan Chockalingam
Bio: Ananthanarayanan Chockalingam is a Professor in the Department of Electrical Communication Engineering, Indian Institute of Science, Bangalore, India, working in the area of wireless communications. He has made pioneering contributions in the area of large-scale/massive MIMO systems. He has served as an Associate Editor of the IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, as an Editor of the IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, and as a Guest Editor for the IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS and IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING. He is a Fellow of the Indian National Academy of Engineering, the National Academy of Sciences, India, the Indian National Science Academy, and the Indian Academy of Sciences. He is a recipient of the prestigious J. C Bose Fellowship from the Science and Engineering Research Board, Department of Science and Technology, Government of India. He is an author of the book on `Large MIMO Systems’ published by Cambridge University Press.
Robert Calderbank
Bio: Robert Calderbank directs the Rhodes Information initiative at Duke University, where he is a Distinguished Professor of Electrical and Computer Engineering, Computer Science and Mathematics. He started his career in the Mathematical Sciences Research Center at Bell Labs, and he left AT&T in 2003 as Vice President for Research. Dr. Calderbank directed the Program in Applied and Computational Mathematics at Princeton University before joining Duke University in 2010. He was elected to the National Academy of Engineering in 2005, and to the American Academy of Arts and Sciences in 2022. Dr. Calderbank received the 2015 Hamming Medal, and the 2015 Shannon Award.
Dr. Calderbank is known for contributions to voiceband modem technology at the dawn of the internet, and for contributions to wireless communication that are incorporated in billions of cell phones. He has also made contributions to quantum error correction that provide a foundation for fault tolerant quantum computation.