Vanessa Chen wins CAREER award for real-time machine learning and cybersecurity

Posted: 

vanessa.chen_.career.2smaller.jpg
Vanessa Chen
In a world where technology is advancing faster than security, changes are on the horizon.

At The Ohio State University, electrical and computer engineering assistant professor Vanessa Chen said smart technology is poised for another breakthrough. Her theory is to convert data into meaningful information and protect it from cyberattack in real time.

One day, every device must automatically be capable of performing its own cybersecurity and machine learning to support big data analysis. Chen said it is making the protection of internet-connected systems, including hardware, software and data, from cyberattacks detrimental.

The National Science Foundation just named Chen among the prestigious 2019 Faculty Early Career Development (CAREER) award winners. Her research proposal, “Bio-Inspired Sensory Interfaces Incorporating Embedded Classification and Encryption,” earned $500,000 in new funding over the next five years from its Electrical, Communications and Cyber Systems (ECCS) program.

“Ubiquitous sensing and computing, leading to rapid growth of big data analysis, will potentially transform the world,” Chen said. “Now, people are trying to turn everything into smart devices.”

The motivation to provide such real-time security within every smart device, she said, is because each one is under threat of malicious attacks by transmitting unencrypted data. Applications sending steady information for continuous health or weather monitoring, for example, are particularly vulnerable.

"The attacker may be able to record the confidential and private information or change the results to broadcast a fake national disaster alert. So, it’s critical to protect the wireless data," Chen said.

The technological drawback for this, however, is the energy required to perform complex machine learning and encryption algorithms.

“It’s hard to use the energy from the environment to power the device, because it would require a large and stable power source like a battery,” Chen said.

Instead, she is working to develop a more energy-efficient circuit architecture to embed into energy-constrained edge devices, performing classification and encryption. An edge device is any piece of hardware that controls data flow at the boundary between two networks, such as routers, routing switches, integrated access devices (IADs), multiplexers, metropolitan area networks (MAN) and wide area networks (WAN). 

"We can have a sensor that can extract and encrypt critical features in situ and then only send low-volume ciphered messages to the central device, so the transmission energy can be highly decreased to enable continuous monitoring," she said.

As the director of the Energy-Efficient Circuits and Systems Lab at Ohio State Chen mentors roughly half a dozen graduate students in the realms of low-power cognitive interfaces for world-to-information computing. Lab work spans the design of wireless transceivers, analog neural networks as well as hardware-based cybersecurity.

One of her students, Jack Hsueh, is focused on low-power and secure sensory interfaces for next-generation Internet-of-Things (IoT) devices. He became the first Ohio State student to win the prestigious ISSCC Analog Devices Outstanding Student Designer Award in 2018.

From a design standpoint, Chen's CAREER research embraces machine learning and cybersecurity through the concept of randomness. 

According to the proposal abstract, data is automatically classified and encrypted within the sensors, changed unpredictably into deterministic noise for transmission. 

“The pipeline chaotic system can be trained with time-varying maps to enhance the strength of the security without creating observable patterns to counter side-channel attacks,” Chen said. "This ensures data integrity and basic authentication for multi-layer security schemes from the edge sensors to the cloud while classification algorithms are performed locally in sensors to achieve rapid analysis and data reduction for wireless communications.”

The transmitted data from the device becomes unclonable, she saids, ensuring complete security. 

Electrical and Computer Engineering Department Assisatnt Professor Nima Ghalichechian and his team also recently won an NSF CAREER award for their work in millimeter wave communication materials

Article by Ryan Horns, ECE/IMR Communications Specialist