2022
Author: Jitian Zhang
Abstract:
Frequency Modulated Continuous Wave (FMCW) radar has become a key sensing modality to enable automotive driving, owing to its ability to function in all-day and all-weather conditions. However, with the increasing number of cars and mounted sensors, interference between radars will impede their ability to detect objects in the environment. In the context of automotive driving, such radar detection errors can be harmful. In this project, we design and implement a time-hopping based radar system and signal processing pipeline to mitigate automotive radar interference. By randomizing the idle time between chirps in a given frame, the power of the interferer can be spread out, while the power of the true reflections add up constructively. This allows us to distinguish interferers from true reflectors. In this work, we formalize this approach and present simulation and experimental results to evaluate our approach.
Author: Waleed Ahmed
Abstract:
Millimeter wave (mmWave) radars are becoming a more popular sensing modality in self-driving cars due to their favorable characteristics in adverse weather. Yet, they currently lack sufficient spatial resolution for semantic scene understanding. In this thesis, we present Radatron, a system capable of accurate object detection using mmWave radar as a stand-alone sensor. To enable Radatron, we introduce a first-of-its-kind, high resolution automotive radar dataset collected with a cascaded MIMO (Multiple Input Multiple Output) radar. Our radar achieves 5cm range resolution and 1.2 degrees angular resolution, 10x finer than other publicly available datasets. We also develop a novel hybrid radar processing and deep learning approach to achieve high vehicle detection accuracy. We train and extensively evaluate Radatron to show it achieves 92.6% AP50 and 56.3% AP75 accuracy in 2D bounding box detection, an 8% and 15.9% improvement over prior art respectively.
Author: Ruihao Yao
Abstract:
Emerging 5G mobile networks are envisioned to be the future of IoT and edge-computing environments. 5G networks feature lower latency, higher capacity and increased bandwidth compared to previous standards. These network improvements will have foreseeable impacts on how people live, work and play. 5G has been proposed and deployed worldwide by many commercial companies. While 5G protocols and working mechanisms are public, the actual implementations still mostly remain closed-source. This thesis reprograms a state-of-the-art LTE simulator to be compatible with 5G network environments. We present a concrete prototype implementation of carrier aggregation and other important 5G features including 5G frequency ranges, femto and pico cell realization and Multicast and Broadcast Service (MBS). We also present experimental results and evaluation of the most common 5G scheduling algorithms’ performances in terms of delay, throughput and latency.
2020
Author: Waleed Ahmed
Abstract:
A recent development integrating Internet-of-Things (IoT) sensing techniques with active noise cancellation (ANC) has demonstrated certain benefits over the conventional methods for ANC, including wideband cancellation without blocking the ear and non-causal adaptive filtering. These benefits, however, can only be observed in acoustic environments with a single noise source. This thesis presents a new design for an IoT-based active noise cancellation system that can effectively cancel multiple independent noise sources. By incorporating multiple reference microphone inputs, the new system can estimate the unique acoustic channels between different sources of noise and the listener. Through simulation and hardware experiments, this new design is evaluated and shown to achieve significant improvement in cancellation over the previous implementation of IoT-based ANC.
2018
Author: Anadi Chaman
Abstract:
This thesis explores the possibility of detecting the hidden presence of wireless eavesdroppers. Such eavesdroppers employ passive receivers that only listen and never transmit any signals, making them very hard to detect. We show that even passive receivers leak RF signals on the wireless medium. This RF leakage, however, is extremely weak and buried under noise and other transmitted signals that can be 3-5 orders of magnitude larger. Hence, it is missed by today’s radios. We design and build Ghostbuster, the first device that can reliably extract this leakage, even when it is buried under ongoing transmissions, in order to detect the hidden presence of eavesdroppers. Ghostbuster does not require any modifications to current transmitters and receivers, and can accurately detect the eavesdropper in the presence of ongoing transmissions. Empirical results show that Ghostbuster can detect eavesdroppers with more than 95% accuracy up to 5 meters away.