Department of Electrical and Computer Engineering

Zhuo Feng, professor in the Department of Electrical and Computer Engineering, recently received a grant worth $599,998 from the National Science Foundation. The project is titled “High-Performance Incremental Spectral Algorithms for Efficient Modeling and Simulation of Large-Scale Integrated Circuits.”

This research plan aims to investigate scalable algorithms for incremental spectral sparsification of large graphs and integrated circuit (IC) networks. The developed framework will allow for efficiently handling large networks that experience streaming edge insertions/deletions in sublinear time, and significantly accelerating SPICE-accurate modeling and simulations of large circuit designs that undergo frequent updates. This research plan will also investigate efficient approaches for incremental modeling and simulation of integrated circuits leveraging incremental spectral sparsification and involve close collaborations with leading industrial experts to evaluate the proposed research.

Feng and two Ph.D. student researchers will investigate a novel algorithmic framework called inGRASS that helps simplify large, complex networks. They will work with leading experts in circuit simulation at NXP Semiconductor to evaluate the research, using realistic circuit designs and comparing the performance against state-of-the-art commercial software and NXP's in-house computer-aided design tools. The resulting approach is intended to help manage changes in the circuit, such as adding or removing connections, much more quickly.

The success of the proposed research will potentially contribute to the advancements in spectral graph theory and electronic design automation (EDA), leading to the development of highly scalable incremental algorithms for solving partial differential equations, sparse matrix problems, spectral graph partitioning and design automation of large-scale ICs.

Hongbin Li, Charles and Rosanna Batchelor Memorial Chair Professor in the Department of Electrical and Computer Engineering, recently received a $599,986 grant from the National Science Foundation for his project, “Look Around the Corner: Ubiquitous RF Sensing with Smart Metasurfaces.” The three-year project seeks to develop an alternative distributed radio frequency (RF) sensing paradigm that can effectively “look” around a corner by leveraging reconfigurable intelligent surfaces (RIS).

Radio frequency sensing is pervasive due to wireless communications, but most RF sensors are unable to locate targets outside of their visual line of sight. A common approach involves deploying multiple RF transmitters and receivers across a surveillance area, but such a system is bulky, expensive and environmentally unfriendly.

Li’s proposed system would overcome these shortcomings by employing RIS: a thin, flat structure comprising numerous small, low-cost metamaterial elements (or materials artificially engineered to have certain properties) that can be independently adjusted to control the reflection of incident RF signals. Like wallpaper, RIS can cover parts of buildings, walls and ceilings, allowing RF engineers to proactively customize the radio environment based on specific needs.

Electrical and computer engineering Anson Wood Burchard Chair Professor Lei Wu and Systems and Enterprises assistant professors Philip Odonkor and Steven Hoffenson have been awarded $590,000 by the National Science Foundation to design and implement a physical computing testbed that will foster problem-based, collaborative learning of cyber-physical systems (CPS) concepts. Cyber-physical systems combine hardware and software components into a network that controls or interacts with a physical environment.

In collaboration with the University of Cincinnati, the three-year project will fill a key gap in CPS education by developing a collaborative tool for learning the multiple complex concepts required for CPS mastery, while fostering a social approach to CPS education. The team will create a comprehensive CPS concept inventory and classification of threshold concepts for optimal CPS training and design a testbed that allows groups of students to work together and interact with cyber-physical systems using a tangible user interface.

The testbed will empower undergraduate students to engage in CPS learning through collaborative play, and the knowledge generated will benefit industry, defense and critical infrastructure by cultivating a competitive U.S. workforce with competence in cyber-physical systems.

Stevens Institute of Technology served as the inaugural university host of the Institute of Electrical and Electronics Engineers (IEEE) INFOCOM Conference for 2023.

The three-day event highlights both theoretical and systems research, drawing researchers from 39 countries to present and exchange innovative contributions and breakthrough ideas in the field of computer networking and closely related areas. Held on the Stevens campus May 17 through 19, the conference was chaired by Stevens President Nariman Farvardin and Department of Electrical and Computer Engineering Chair Min Song.

The INFOCOM 2023 program included 63 technical paper sessions, 20 workshops, a student poster session, a demo session and panel discussions. The keynote address was delivered by Ion Stoica, a professor in the Computer Science Division at the University of California, Berkeley, and director of UC Berkeley’s Sky Computing Lab.

After being held virtually for the past three years, the 42nd annual conference convened in hybrid mode, with programs taking place in person and streamed live to remote attendees via Zoom.