Research Projects
Sponsor: National Science Foundation (NSF)
Research Topic: Learning Over Ultra-Dense Wireless Network
(10/2020-09/2024)
Synopsis of the Project
An accurate characterization of the statistical behavior of wireless networks is crucial in the analysis, design, and deployment of real-world wireless networks. In the past decade, point processes without spatial repulsion, such as Poisson point processes, have been intensively applied to model and analyze the performances of wireless networks. However, these point processes may not be suitable for modeling and analyzing real-world wireless networks with diverse types of spatial repulsion. This project proposes two families of point processes generalizing various known repulsive processes, which are able to accurately characterize real-world repulsive phenomena in a wireless network such as non-linear and/or asymmetric repulsion. In addition, contrary to the existing results in the literature that are mostly semi-analytical or numerical, the project aims at explicit closed-form characterizations by advanced tools from stochastic geometry and random matrix theory. The results in this project will provide key insights and a new benchmark in the design of various wireless networks.
The proposed research resides in the interdisciplinary area of stochastic geometry and wireless networks. The considered point processes are able to characterize diverse nodal repulsion phenomena in ultra-dense wireless networks so as to fundamentally improve the existing modeling and analysis framework of wireless networks with spatial randomness where the impact of node repulsion is ignored. The project has also been motivated by the fact that state-of-art analytical tools in the mathematics community are far from being fully utilized in the wireless networking community. The project aims at the analytical characterization of performance metrics including user association statistics, interference statistics, link rate, and distributed learning, which are based on the closed-form evaluations of some fundamental performance measures. Among other consequences, the proposed research will lead to new scaling laws useful in the deployment of ultra-dense wireless networks for practitioners. The synergy of the PIs brings about the latest ideas and approaches from stochastic geometry and random matrix theory aiming at new breakthroughs in understanding the fundamental behavior of wireless networks. The outcome of the proposed research also has applications in other domains, such as in machine learning and data science, where the project results offer new algorithms for sampling, marginalization, conditioning, and other inference tasks.
Personnel
Principle Investigator: Dr. Chun-Hung Liu
Graduate Research Assistant: Ms. Roza J. Farahani
Academic Collaborator and Industry Partner
Dr. Lu Wei with the Department of Computer Science at Texas Tech University
Dr. Ivy Kelley at C Spire
Broad Impacts: Conference, Outreach, and Educational Activities
Tutorial at the IEEE VTC Fall 2021: Intelligent Spectrum Access and Sharing in Wireless Networks with Reconfigurable Reflecting Surfaces
ECE's Research Symposium, Oct. 22, 2021: Research Presentation for the NSF Project (See the link here)
STEM Days, Aug. 26-27, 2021: Lab Tour and Research Demo for High School Students (See the link here)
Computing Workshop, Jun. 1, 2022: Workshop for High-School Teachers (See the link here)
Workshop of Learning over Networks, Aug. 16, 2023: Research Demonstrations for High-School Teachers and Students (see link here)
Educational Activities:
Courses offered: (i) Optimization for Machine Learning (Fall 2022, Spring 2024) (ii) Machine Learning for Wireless Networks (Spring 2023)
Related Publications
C.-H. Liu and M. Gamarra, "Optimal Client Sampling for Wireless Federated Learning With Random Client Participation", submitted to IEEE International Conf. on Commun. (ICC), 2024.
F. Freitag, L. Wei, C.-H. Liu, Me. Selimi, and L. Veiga, "Server-side Adaptive Federated Learning over Wireless Mesh Network", in Proc. of Int. Conf. on Information Technology & Systems, 2023.
C.-H. Liu, K.-T. Feng, L. Wei, and Y. Luo, "Spatio-Temporal Federated Learning for Massive Wireless Edge Networks", In Proc. of IEEE Int. Conf. on Commun. (ICC), 2022.
F. Freitag, P. Vilchez, L. Wei, C.-H. Liu, M. Selimi, and I. Koutsopoulos, "Demo: An Experimental Environment Based OnMini-PCs For Federated Learning Research", in Proc. of IEEE 19th Annual Consumer Commun. & Net. Conf. (CCNC), 2022.
C.-H. Liu, K.-T. Feng, L. Wei, and Y. Luo, "Spatio-Temporal Federated Learning for Massive Wireless Edge Networks", In Proc. of IEEE Int. Conf. on Commun. (ICC), 2022.
C.-H. Liu, D.-C. Liang, R-H. Gau, and L. Wei, "Modeling and Analysis of Intermittent Federated Learning Over Cellular-Connected UAV Networks", In Proc. of IEEE Vehicular Technology Conference (VTC), Spring 2022.
Freitag, F., Vilchez, P., Wei, L., Liu, CH., Selimi, M. (2022). "Performance Evaluation of Federated Learning Over Wireless Mesh Networks with Low-Capacity Devices, Information Technology and Systems", ICITS 2022. Lecture Notes in Networks and Systems, vol 414. Springer.
F. Freitag, P. Vilchez, L. Wei, C.-H. Liu, M. Selimi, and I. Koutsopoulos, "Demo: An Experimental Environment Based OnMini-PCs For Federated Learning Research", in Proc. of IEEE 19th Annual Consumer Commun. & Net. Conf. (CCNC), Apr. 2022.
F. Freitag, P. Vilchez, L. Wei, C.-H. Liu, and M. Selimi, "Poster: Testbed in Wireless City Mesh Network with Application to Federated Learning Experiments", in Proc. of the 11th Int. Conf. on the Internet of Things, Nov. 2022.
C.-H. Liu, K.-T. Feng, L. Wei, and Y. Luo, "Spatio-Temporal Federated Learning for Massive Wireless Edge Networks", in Proc. of IEEE Int. Conf. on Commun. (ICC), 2022.
D.-C. Liang, C.-H. Liu, R-H. Gau, and L. Wei, "Modeling and Analysis of Intermittent Federated Learning Over Cellular-Connected UAV Networks", in Proc. of IEEE Vehicular Technology Conference (VTC), Spring 2022.
C.-H. Liu, D.-C. Liang, and R.-H. Gau, "A 3D Modeling Approach to Tractable Analysis in UAV-Enabled Cellular Networks", in Proc. of IEEE Int. Conf. on Commun. (ICC), Jun. 2021.
C.-H. Liu, D.-C. Liang, M. A. Syed, and R.-H. Gau, "A 3D Tractable Model for UAV-Enabled Cellular Networks with Multiple Antennas", IEEE Trans. on Wireless Communications, vol. 20, no. 6, pp. 3538-3554, Jun. 2021.
Analysis and Adversarial Learning of High-Dimensional Noisy Data for Power Electronics Systems
Sponsors: NSF
ECCS-EPCN: Degradation-Aware Self-Healing Control of Power Electronics Systems (8/16/2022-8/15/2025)
Project Summary
Computational power is everywhere. Sensors are increasingly low-cost and ubiquitous. Despite the extensive resources, modern power electronics systems (PESs) cannot pinpoint its degradation status and, hence, cannot perform self-healing to prevent costly failures. For example, wind turbines or photovoltaic (PV) systems are subjected to extreme temperature and humidity swings from -30ºC to 55ºC and 30% to 100% (e.g., offshore applications). Such a harsh climate and thermal (C&T) swings rapidly increase the failure rate and maintenance costs by up to 30% of the overall generation cost. Suppose their degradation and, hence, remaining useful lifetime (RUL) can be accurately measured or precisely predicted in advance. In that case, we can utilize existing PESs software or hardware to perform proactive self-healing through the adaptive control of degradation evolution, accumulation, acceleration, and, hence, RUL changes of the building blocks of power electronics in increasingly complicated modern energy systems. This could substantially enhance reliability, scheduling flexibility, and controllability while preventing costly downtimes. The outcome of this project will be utilized for interactive and hands-on learning programs to inspire K-12 children’s interest in STEM fields.
This project will model the degradation of wide bandgap (WB) power switches under real-world C&T swings, which poses the critical bottleneck of exploiting degradation-aware self-healing (DASH) control in modern PESs. Specifically, we will develop a cascade generative adversarial networks learning and data purification strategy to effectively model the large reliability data of power electronics under a real-world C&T condition. The formulated data-driven models and multi-sensory tools will be fundamentally more accurate than state-of-the-art. Moreover, we will develop a systematic DASH control framework, enabling lifetime managed PES operations by understanding four system health conditions (healthy, intermediate degradation, self-healing, and failure) instead of a traditional heuristic assumption (healthy and failure). The formulated RUL estimation and DASH control tools are able to fundamentally transform the current design and control practices, creating a seamless integration of reliable WB switches into the wide spectrum of power electronics and energy systems under diverse C&T conditions. This will accelerate the migration toward an energy-efficient grid and transportation electrification while minimizing development cost and period and preventing unplanned downtime and catastrophic failures.
Typical Related Publications
S. A. Jawdeh, S. Choi, and C.-H. Liu, "Model-Based Deep Learning for Cyber-Attack Detection in Electric Drive Systems", in Proc. of IEEE Applied Power Electronics Conf. (APEC), 2022.
J.-Y. Wu, L.-C. Huang, W.-H. Li, C.-H. Liu, and R.-H. Gau, "Greedier Is Better: Selecting Multiple Neighbors per Iteration for Sparse Subspace Clustering", submitted to IEEE Trans. on Information Theory, Apr. 2022.
J.-Y. Wu, L.-C. Huang, M.-H. Yang, and C.-H. Liu, "Sparse Subspace Clustering via Two-Step Reweighted L1-Minimization: Algorithm and Provable Neighbor Recovery Rates", IEEE Trans. on Information Theory, vol. 67, no. 1216-1263, pp. 2814-2831, Feb. 2021.
J.-Y. Wu, W.-H. Lee, L.-C. Huang, Y.-P. Lin, C.-H. Liu, and R.-H. Gau, "Provable Noisy Sparse Subspace Clustering using Greedy Neighbor Selection: A Coherence-Based Perspective", submitted to IEEE Trans. on Pattern Analysis and Machine Intelligence, Sep. 2021.
Adaptable Spectrum Sensing and Learning for UAV-Enabled Networks with Reconfigurable Intelligent Surfaces
Sponsors: Air Force Research Laboratory (AFRL) (08/01/2022-07/31/2025)
Project Summary
Efficiently utilizing the radio spectrum is an imperative goal for future wireless networks that need to widely and seamlessly cover various mobile devices distributed over a large territory. Achieving this goal is a big challenge due to the uncontrollable random nature of wireless channels, which makes transmitted signals hard to be effectively detected at anytime and anywhere. Such a challenge becomes even more severe when different kinds of aerial and ground mobile devices using multiple radio access technologies coexist in a wireless network. The recent metamaterial technology of developing intelligent reflecting surfaces (IRSs) provides a feasible solution to alleviating the uncontrollability of wireless propagation environments such that the random characteristics of wireless channels are no longer completely uncontrollable, thereby letting wireless networks able to achieve wider and more uniform coverage. The scope (goal) of this project aims to investigate novel joint data-driven and model-driven (a.k.a, hybrid-driven) methodologies of Spectrum Sensing and Sharing (SSS) in wireless networks with IRSs. We will propose new and effective SSS solutions that not only make mobile devices jointly conduct spectrum sensing and access in wireless networks, but also let them have self-learning capability of quickly adapting to wireless environmental changes so as to improve the spectral utilization of future wireless networks that can provide different kinds of mobile devices, such as ground mobile handsets, low-and-high-altitude platforms (such as unmanned aerial vehicles), low-earth-orbit (LEO) satellites.
Typical Related Publications
C.-H. Liu, M. D. Syed, and L. Wei, "Toward Ubiquitous and Flexible Coverage of UAV-IRS-Assisted NOMA Networks", In Proc. of IEEE Wireless Commun. and Net. Conf. (WCNC), 2022.
Federated Deep Reinforcement Learning for Security in Wireless Cyber-Physical Systems
Sponsors: AFRL/DHS
(02/01/2022-01/31/2025)
Project Summary
Safety is of paramount importance to autonomous applications which need real-time and delay-sensitive data transmission over wireless networks. A promising approach to timely and optimally fulfilling such low-latency autonomous applications is to adopt mission-critical communications integrated with learning-based edge computing. The objective of this project is to attain mission-critical intelligence acquisition for wireless cyber-physical systems (CPS). In this project, we aim to propose new Federated Learning (FL) algorithms with intermittent updates due to channel fading and develop a tractable and network-wide framework of analyzing the algorithms for mission-critical communications that can be used to study the learning capability, security, and privacy performances of FL in delay-sensitive scenarios. Exploiting the limits of these performances will not only reveal how communication reliability and FL couple each other, but also provide insights into how these limits can be relaxed by optimizing the settings of wireless CPS with latency constraints. The research outcomes of this project are expected to not only macroscopically reveal how the performances of the proposed FL are influenced by all communication activities in a wireless network, but also have a significant and positive impact on tackling the problem of how to optimally deploy and manage a wireless CPS in order to make FL over wireless networks able to swiftly learn with privacy and security.
Typical Related Publications
C.-H. Liu, K.-T. Feng, L. Wei, and Y. Luo, "Spatio-Temporal Federated Learning for Massive Wireless Edge Networks", In Proc. of IEEE Int. Conf. on Commun. (ICC), 2022.
C.-H. Liu, D.-C. Liang, R-H. Gau, and L. Wei, "Modeling and Analysis of Intermittent Federated Learning Over Cellular-Connected UAV Networks", In Proc. of IEEE Vehicular Technology Conference (VTC), Spring 2022.
K.-C. Chen, S.-C. Lin, J.-H. Hsiao, C.-H. Liu, A. F. Molisch, and G. Fettweis, "Wireless Networked Multi-Robot Systems in Smart Factories", Proceedings of the IEEE, vol. 109, no. 4, pp. 468-494, Apr. 2021.
C.-H. Liu, D.-C. Liang, K.-C. Chen, and R.-H. Gau, "Ultra-Reliable and Low-Latency Communications Using Proactive Multi-cell Association", IEEE Trans. on Communications, vol. 69, no. 6, pp. 3879-3897, Jun. 2021.