A recent report by the US Department of Energy defines the area of scientific machine learning as “a core component of artificial intelligence (AI) and a computational technology that can be trained, with scientific data, to augment or automate human skills”, which has “the potential to transform science and energy research”. We explore the potential of scientific machine learning methods to problems in computational electromagnetics starting from standard microwave structure design and multiphysics modeling, employing an unsupervised learning strategy based on Physics-Informed Neural Networks (PINN). PINNs directly integrate physical laws into their loss function, so that the training process does not rely on the generation of ground truth data from a large number of simulations (as in typical neural networks). Moreover, we demonstrate the impact of machine learning on the computational modeling of radiowave propagation scenarios. We build convolutional neural network models that can process the geometry of indoor environments, along with physics-inspired parameters, to rapidly estimate received signal strength (RSS) maps. We show the *generalizability* of these models, which is their ability to "learn" the physics of radiowave propagation and produce accurate modeling predictions in new geometries well beyond those included in their training set. These models can be used to rapidly optimize the position of transmitters in wireless area networks, to maximize coverage or other relevant metrics. Co-sponsored by: STARaCom Speaker(s): Costas Sarris Room: Room 603, Bldg: McConnell Eng. Building , McGill Unversity, Montreal, Quebec, Canada

This is a weekly session of the CIT Summer Series, with Dr. Rami Abielmona presenting Distilling AI: The Hitchhiker's Guide : In this talk, I will present real-world AI/ML Big Data solutions involving unique learning algorithms that allow one to process vast amounts of critical information combined with knowledge acquired from specific domains. These unique models and architectures continually deliver the most accurate information possible in order to constantly optimize the decision maker’s domain awareness. Attendees will learn innovative concepts such as the five levels of Big Data Analytics, the various variations of AI including Machine Learning, Deep Learning and Machine Intelligence, the transformational changes that AI can bring about in the near future as well as the challenges and opportunities to deploy AI-ready applications in mission-critical tactical environments. Speaker(s): Dr. Rami Abielmona, Virtual: https://events.vtools.ieee.org/m/363999

Microwave photonics (MWP) is a typical optical signal processing application for optical communications, antenna systems, and 5G/6G networks. At the same time, optical frequency combs (OFC) and programmable optical filters enable this system to be reconfigurable. There are several approaches to creating OFC lines, such as the micro-ring resonator, cascaded electro-optic modulator, and mode-locked laser (MLL), in which the quantum dash (QDash) MLL is an ideal on-chip OFC source to provide low relative intensity noise (RIN), narrow linewidth, and flat comb spectrum. In this seminar, we will present three typical applications of MWP systems using QDash MLL as the OFC source. The photonic beamforming illustrates a phased antenna array system that can do directional radiation and scanning. The MWP filter is a reconfigurable finite impulse response (FIR) filter, and a specially designed MWP filter can also be used for instantaneous frequency measurement. In partnership with the (https://nrc.canada.ca/en/research-development/research-collaboration/programs/high-throughput-secure-networks-challenge-program)Challenge program at National Research Council (NRC), we invite you to join this virtual seminar series to promote scientific information sharing, discussions, and interactions between researchers. Co-sponsored by: National Research Council, Canada Speaker(s): Yuxuan Xie , Virtual: https://events.vtools.ieee.org/m/366426

This is a weekly session of the CIT Summer Series, with Nael Abu-Ghazaleh presenting Security challenges and opportunities at the Intersection of Architecture and ML/AI : Machine learning is an increasingly important computational workload as data-driven deep learning models are becoming increasingly important in a wide range of application spaces. Computer systems, from the architecture up, have been impacted by ML in two primary directions: (1) ML is an increasingly important computing workload, with new accelerators and systems targeted to support both training and inference at scale; and (2) ML supporting architecture decisions, with new machine learning based algorithms controlling systems to optimize their performance, reliability and robustness. In this talk, I will explore the intersection of security, ML and architecture, identifying both security challenges and opportunities. Machine learning systems are vulnerable to new attacks including adversarial attacks crafted to fool a classifier to the attacker’s advantage, membership inference attacks attempting to compromise the privacy of the training data, and model extraction attacks seeking to recover the hyperparameters of a (secret) model. Architecture can be a target of these attacks when supporting ML, but also provides an opportunity to develop defenses against them, which I will illustrate with three examples from our recent work. First, I show how ML based hardware malware detectors can be attacked with adversarial perturbations to the Malware and how we can develop detectors that resist these attacks. Second, I will also show an example of a microarchitectural side channel attacks that can be used to extract the secret parameters of a neural network and potential defenses against it. Finally, I will also discuss how architecture can be used to make ML more robust against adversarial and membership inference attacks using the idea of approximate computing. I will conclude with describing some other potential open problems. Speaker(s): Nael Abu-Ghazaleh, Virtual: https://events.vtools.ieee.org/m/364001

Abstract: Renewable energy systems are gaining increasing importance throughout the world. Proliferation of renewable energy sources in power systems has led to new opportunities and challenges in control, stability, protection, power quality and operation of power systems. Power electronics is an enabling technology for effective integration of renewables into the electrical grid. The aim of this talk is to provide an overview of a range of important power electronic applications in microgrids and smart grids with a high share of renewables. Participants will mainly gain insight into the control of power electronic converters in such applications. The main topics are as follows: -Distributed generation and microgrids -Photovoltaic systems -Wind energy systems -Grid codes for renewable energy systems Speaker(s): Prof. Mehdi Savaghebi Virtual: https://events.vtools.ieee.org/m/365502