Ph.D. Students Follow the Late Professor’s Example in Scholarship, Leadership and Service

The electrical and computer engineering community at UVA commemorates the legacy of Professor Malathi Veeraraghavan through a graduate fellowship in her name. The Department established the merit-based, donor-funded fellowship program in 2021. The 2022-23 fellows are Yujia Mu, a Ph.D. student of computer engineering, and Farjana Tonni, a Ph.D. student of electrical engineering.

“I can think of no professor more committed to diversity and inclusion than Professor Veeraraghavan,” said Scott Acton, chair of the Charles L. Brown Department of Electrical and Computer Engineering. “Malathi was also committed to mentorship and training. She was tireless, even writing mentoring notes and correcting papers while walking her dog! I am certain that Malathi would be proud to support the training of these two women in their journey to become leaders of engineering.”

Mu had the privilege of working directly with Veeraraghavan, who passed away on May 11, 2020. Veeraraghavan’s group members focused on network security, software-defined networks and networks for industrial robotics and smart communities. Mu used machine learning methods to detect unusual network activity, part of a long-term cyber-defense initiative that provided network managers with early warning tools to detect intrusions while preserving data privacy.

“Professor V understood and guided everyone's projects,” Mu recalled. “She took notes in our one-on-one meetings, to memorize and track our progress. She was a responsible tutor who cared about her students, actively communicating with us about our studies and about life.”

Mu began her graduate study at the UVA School of Engineering and Applied Science in 2017, earning her Master of Engineering degree in computer engineering in 2019. While in the job market, Mu sought to join Veeraraghavan’s research group as a temporary research assistant. But Veeraraghavan had bigger plans; she encouraged Mu to continue her research in the Ph.D. program.

In Veeraraghavan’s group, Mu developed machine learning algorithms and models for next-generation networking, designing a deep learning architecture to reliably and simultaneously classify streamed data files to many end-users, a network function called multi-casting. Mu proposed a quality-of-service-driven data classification pipeline to achieve the optimal file transmission rate based on the type, volume and quality of data distributed to through a network at any point in time.

While still early in her Ph.D. program, Mu was co-advised by Veeraraghavan and Cong Shen, assistant professor of electrical and computer engineering. She is a member of Shen’s laboratory for intelligent communication and networking.

Cong Shen and Yujia Mu work with low-power devices in lab

Ph.D. student Yujia Mu (right) inspects a NVIDIA Jetson while Cong Shen, assistant professor of electrical and computer engineering, explains how to apply federated learning to this and other low-power devices.

Mu specializes in a machine learning technique that is ideal for data privacy and edge computing called federated learning, which she applies to wireless communication networks. For example, Mu proposed a method for federated learning tasks when the quality of the channel between the transmitter and receiver is poor, such that an originally successful communication may suddenly become disconnected due to dynamics in the wireless medium. Mu’s solution to this problem, termed deep fading, was accepted by the prestigious IEEE International Symposium on Information Theory, colloquially known as ISIT 2022.

Mu is exploring a sub-area called federated split learning that achieves the predictive power and other benefits of machine learning without having to share data between devices and network servers. The deep neural network — the element that teaches computers how to process data like a human brain, via interconnected nodes of a layered, mathematical structure — is divided into two parts. One part is trained on users’ devices while the other is trained on the network server.

Professor Veeraraghan’s legacy and the fellowship established in her name encourages students to be well-rounded and explore interests outside academe. For Mu, that passion lives on the basketball court, a talent she developed at an early age. She trained every day in her high school and played in China’s NCAA-level basketball league for Tsinghua University in Beijing, where she earned her Bachelor of Engineering degree in electrical engineering.

  • Yujia Mu shoots three-pointer on the basketball court

    Ph.D. student Yujia Mu plays in the women’s league of the North East China Basketball Association, which UVA recently joined.

  • Yujia Mu and women's basketball teammates

    Mu captained UVA’s team in the league’s April 2022 tournament, competing against 10 other college teams.

  • Yujia Mu completes a lay-up on the basketball court

    She averaged 16 points per game; in the last game, she made four three-pointers and scored 22 points in total.

When Mu came to the United States, she was determined to continue in her sport, playing basketball on men’s teams. Nowadays, Mu plays in the women’s league of the North East China Basketball Association, which UVA recently joined. Mu captained UVA’s team in the league’s April 2022 tournament, competing against 10 other college teams. She averaged 16 points per game; in the last game, she made four three-pointers and scored 22 points in total.

“I think that Professor V. would be proud of how I have carried the persistence, teamwork and spirit of the basketball court into my research,” Mu said.

Tonni, who did not have the opportunity to meet Professor Veeraraghavan, found other strong, female leaders in her advisor, associate professor Mona Zebarjadi, and Jennifer L. West, UVA Engineering’s dean.

“Seeing other female leaders in our field inspires me,” Tonni said. “As an engineer in a STEM field, I see a lot of male-dominated technical sections. Working alongside Professor Zebarjadi and meeting Dean West shows me that we can do much better. We can go up the ladder by our own strength and creativity.”

Tonni earned her master’s degree in electrical and electronic engineering from the University of Dhaka, located in Dhaka, Bangladesh, in 2016. She specializes in materials that possess the potential to change thermal energy into electricity, a conversion process involving charge and heat transport, with electrons and phonons as the primary carriers. Engineers can make the conversion process more efficient by changing a material’s structure to achieve desired electrical and electronic material properties.

Tonni was excited by Zebarjadi’s research in thermoelectrics and semiconductor devices. Zebarjadi, who holds joint appointments in electrical and computer engineering and materials science and engineering, leads the energy science nanotechnology and imagination lab, whose members design materials with altered thermal and electrical properties and new hybrid energy conversion devices.

  • Farjana Tonni in lab

    Ph.D. student Farjana Tonni works with a cryostat that is able to function in the temperature range of 10 Kelvin – 800 Kelvin.

  • Farjana Tonni and Prof. Zebarjadi work side-by-side in the lab

    Ph.D. student Farjana Tonni (left) and Mona Zebarjadi, asociate professor of electrical and computer engineering and materials science and engineering, adjust the focus of the microscope on a packaged sample placed in a chamber with an optical window on top for thermal imaging at an ultra-low temperature.

  • Farjana Tonni takes a thermal image of a thin film

    Ph.D. student Farjana Tonni takes a thermal image of a nano-structured silicon thin film device at room temperature using the Microsanj thermo-reflectance imaging system. This 2D temperature map of the device under bias helps to study heat transport in transient and steady-state and analyze thermal management in the device.

“We are looking into applications for wearable devices,” Tonni said. “Imagine a backpack made of a material that can efficiently convert your body heat into electronic heat to charge your cell phone.”

Tonni’s research will identify specific characteristics of 2D materials and thin films with unique electronic, mechanical, thermal and optoelectronic properties that make them efficient conductors of heat and electricity. Tonni’s goal is to fabricate and characterize the performance of these materials for thermoelectric energy conversation applications.

Tonni is learning how to fabricate 2D materials in UVA’s newly renovated and expanded clean room, the Innovations in Fabrication or IFAB facility, so she can measure characteristics of material samples she has made herself.

Tonni would have found a kindred spirit in Veeraraghavan, remembered for her multicultural, global outlook and passion for teaching. Tonni was born and grew up in Bangladesh, and actively supports UVA’s community of Bangladeshi students. She served as general secretary and remains involved in an informal, student-run committee that welcomes Bangladeshi students to Grounds, helping them become familiar with Charlottesville and the University.

The committee also leads Bangladeshi students’ annual celebration of International Mother Language Day, proclaimed by the United Nations and its Educational, Scientific and Cultural Organization to promote unity in diversity and international understanding through multilingualism and multiculturalism. While raising funds for International Mother Language Day events at UVA, Tonni had an opportunity to show off her culinary skills, selling her home-cooked bites and treats with Charlottesvile Tom Tom Festival goers.

Tonni's concern for others extends well beyond her home country’s borders. She joined Zebarjadi in STEM programs at local middle schools and high schools, demonstrating lab skills and conducting small experiments, and hopes to continue this outreach in the next school year. Looking to the future, Tonni aims to strengthen the connection between her service and academics, with contributions to engineering societies and their student-led chapters.