Research Overview


Yang (Cindy) Yi, Ph.D., is an assistant professor in the Bradley Department of Electrical and Computer Engineering (ECE) at Virginia Tech (VT). Her research is primarily focused on Very Large Scale Integrated (VLSI) Circuit and Systems, Neuromorphic Computing, Congitive Computing and Communications, Emerging Technologies, Hardware Reliability and Variability. Her major accomplishments include designing and fabricating analog neural chips for spiking recurrent neural network, designing and analyzing the energy-efficient circuits for green computing and communication system, and exploring the application of recurrent neural network and machine learning to wireless communications and cybersecurity. Her research has been funded by National Science Foundation (NSF), Department of Defense (DoD), NSF Experimental Program to Stimulate Competitive Research (EPSCoR), Air Force Office of Scientific Research (AFSOR), Air Force Research Lab (AFRL), and Industrial companies. Some of her current research projects are listed as follows:

Neuromorphic Electronic Circuits Design and Automation for Brain-inspired Computing Systems

Neuromorphic computing systems, which represent a type of non-traditional architecture that encompasses evolutionary systems, hold great promise for leveraging these behaviors to address specific classes of mission-critical problems that have not been solved by current state-of-the-art CMOS digital computing. Reservoir computing is a recently developed machine-learning paradigm whose processing capabilities rely on the dynamical behavior of recurrent neural networks. The goal of this project is to build a new class of computationally efficient spike timing-dependent encoders and delay-based reservoirs within reservoir networks. These advances could offer potentially disruptive capabilities in real-time signature analysis, time-series predictions, and environmental perception for autonomous operations and dynamic-control systems. This work will result in an agile analog/mixed signal integrated-circuit implementation of a spike-time encoding circuit as a signal conditioner and an electronic reservoir as a dynamic processor for reservoir computing systems. Based on the close collaboration with the AFRL, we successfully fabricated and tested the analog spiking time dependent encoder using Global Foundries CMOS 180nm technology. To the best of our knowledge, the introduced neuron circuit is the world’s first to present the sensory data using inter-spike interval temporal encoding.


Microscope image of the fabricated spiking neural chip in 2015 and 2016

Energy Efficient and High Performance Three Dimensional IC Design for Green Computing and Renewable Energy System

To combat the global energy crisis and reduce the negative environmental impacts (e.g. global warming), green computing and renewable energy (e.g. solar energy) have attracted attention from both academia and industry recently. Energy harvesting integrated circuits and systems play important roles in enhancing the efficiency of devices in harvesting energy from environmentally-friendly “green” resources and converting them into electrical energy. Current two-dimensional (2D) integrated circuit technology is approaching its physical and material limits. One promising solution is to integrate circuits in three dimensions, providing higher system speed, higher density, lower power consumption, and smaller footprint. This project provides promising modeling and design solutions for three dimensional (3D) integrated circuits to enable new paradigm for green computing and renewable energy applications, by integrating different technological compartments such as CMOS, nano-devices, logics, memory, and analogue sensors.

Hardware Implementations of Machine Learning and Artificial Intelligence for Applications in Wireless Communications and Cybersecurity.

Reservoir computing is a recent neurologically inspired concept for processing time dependent data that lends itself particularly well to hardware implementation by using the device physics to conduct information processing. The proposed research develops novel and fundamental methodologies for data representation using hardware spike timing dependent encoding for neuromorphic processors; explores the applications of neuromorphic computing in wireless communications including channel estimation in Multiple-Input and Multiple-Output (MIMO) and terahertz communications, as well as spectrum sensing in cognitive radio. This project will introduce an interdisciplinary approach for exploring the application of neuromorphic computing in wireless communications. This will bridge high performance computing, nano technology, and telecommunications, and improve the computational efficiency and accuracy of channel estimation and spectrum sensing. Furthermore, in the realm of cyber physical systems, the tight interaction among physical objects which collect and transmit large volume of data place security threats under the spotlight of attention. With this enormous amount of data that is constantly flowing through the network, it becomes a challenge task for analysts to monitor the huge amount of security-related information that is being exchanged for anomaly detection; that, in addition to the lack of qualified security experts. In order to adequately defend vital systems against potential attacks, RC offers a promising solution to automate many of these security-related tasks, especially with the continuous growth of the flowing data in terms of scale and complexity. We will apply the parallel and scalable reservoir computing networks to build efficient online monitoring and robust decision support modules, and provide cross-linking solutions for anomaly detection.

Integrated Circuit/Transceiver Design and Optimization for Wireless/Cellular Networks, Radio Frequency (RF) Energy Harvesting Systems, and Ehealth Systems.

The RF energy harvesting systems, LTE advanced relay networks, and Ehealth systems promises tremendous potential to improve resource management, increase productivity, and to enable new business models across business and consumer markets. As wireless communication technologies are advancing by leaps and bounds, more and more devices are able to access internet. This project focus on the energy efficient and high performance transceiver design for both RFID and wireless sensor networks (WSN). We successfully designed a full CMOS power amplifier for wireless communication system. Two-transistor cascade structure is adopted as drive stage which has higher gain than one-transistor structure. The common source structure, which has high gain property, is applied on the output stage. The transistor of output stage is carefully designed to get a balance between output power and power added efficiency (PAE). We analyze and optimize the energy efficiency, spectral efficiency, bandwidth of circuit/transvers applied in the wireless networks, RF energy harvesting systems, and Ehealth systems.

Hardware Reliability and Variability Analysis in High Performance Computing Systems

As technology keeps scaling down, hardware variability, such as process variations (PV) and negative bias temperature instability (NBTI), emerges as a growing challenge in the modern GPGPUs (general-purpose computing on graphics processing units). PV induces significant delay variations statically, while NBTI dynamically slows down the GPGPUs. Each computing core (i.e., streaming multiprocessor) in GPGPUs supports thousands of simultaneously active threads, and requires a large register file. Such a sizable register file is very sensitive to the hardware variability, and becomes one of the major units in determining the core frequency. We investigated a set of techniques that mitigate both the PV and NBTI impacts on GPGPUs register file.

Inter-disciplinary research between VLSI and other emerging areas

  • Smart Grid: the development of smart grids involves modernizing the world’s electricity systems to achieve high levels of efficiency, reliability, resiliency, and security. Accordingly, design automation methods and tools need to make significant contributions to advancing the state of the art in hardware design.

  • Wireless Healthcare: wireless healthcare is to enable low power, high accuracy, and reliable communication through designing new communication devices as well as sensing/detection algorithms.

  • Big Data Analytics: with big data analytics, data scientists and others can analyze huge volumes of data that conventional analytics and business intelligence solutions can't touch. We would like to explore the application of reservoir computing in the process of the examining big data to uncover hidden patterns, unknown correlations and other useful information that can be used to make better decisions.

  • Cybersecurity: protecting the information systems from theft or damage to the hardware, software, and disruption or misdirection of the services they provide is very important. We would like to focus on a physical implementation in hardware of neural network algorithms for near- or real-time data mining, sorting, clustering, and segmenting of data to detect and predict criminal behavior.