Seminar: The Jacob Ziv Communication and Information Theory seminar
Extremum Encoding for Distributed Time-Delay Estimation
| Modern distributed sensing systems—ranging from IoT-based localization to ad-hoc sensor networks—are increasingly challenged by stringent communication constraints. A fundamental task in these systems is time-delay estimation (TDE) between spatially separated sensors, which is essential for applications such as emitter localization and synchronization. This talk presents a novel approach to joint data compression and time-delay estimation tailored for such resource-constrained sensing scenarios.
We introduce a simple yet high-performing strategy based on what we term “extremum encoding”, whereby one sensor transmits (to the other) only the index of the maximal observed signal value. Subsequent joint processing of the encoded message with the locally observed, time-delayed, noisy signal gives rise to our proposed time-delay “maximum-index”-based estimator (MIE). We will discuss the theoretical guarantees of this scheme, including its exact error exponent and consistency, and compare its empirical performance against classical and state-of-the-art alternatives. The method is computationally efficient and requires no prior knowledge of the signal-to-noise ratio. Our results not only demonstrate the practical viability of extremum encoding but also suggest a broader paradigm shift in designing inference-oriented compression schemes for distributed systems. |
| Amir Weiss is with the Faculty of Engineering at Bar-Ilan University. He received the B.Sc. (magna cum laude), M.Sc. and Ph.D. degrees in electrical engineering from Tel Aviv University (TAU), Tel-Aviv, Israel, in 2013, 2015 and 2020, respectively. From 2019 to 2020, he was a postdoctoral fellow at the Department of Computer Science and Applied Mathematics, Weizmann Institute of Science, Rehovot, Israel, and a from 2020 to 2023 a postdoctoral associate at the Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge, MA, USA. His main research areas are in statistical and digital signal processing, estimation theory and machine learning. He has won a number of awards for his research, including the 2021 ICASSP Outstanding Paper Award.
|