As a student at Istanbul Technical University in the late 1990s, Melike Erol-Kantarci got her first mobile phone — a clunky Motorola, just voice and text at 2G speed. She also received her calling: “a passion for networks,” inspired by a computer networks course with her soon-to-be PhD advisor.
Today, the professor of engineering holds the Canada Research Chair in AI-enabled Next-Generation Wireless Networks and her lab group is imagining powerful 6G wireless network possibilities, the successor to the 5G wireless network now being rolled out. (“G” stands for “generation.”)
“We are just seeing the tip of the iceberg in possible 5G applications, let alone 6G. With 6G, maybe we’ll have holographic communications,” she says, reflecting on her current pandemic-era online teaching. “As a professor, my hologram would appear in the students’ spaces just as if they were in a classroom.”
Like her subject, Erol-Kantarci’s wireless research is digital: she uses specialized software to build “digital twin” 5G and 6G networks informed with real network data. This includes the millions of algorithms that direct digital traffic at a wireless base station. The algorithms control everything from the routing to scheduling of billions of digital packets of text, voice and data from our myriad wireless devices.
Erol-Kantarci explores how to optimize these algorithms by “training” them, using deep reinforcement learning, a subfield of artificial intelligence. “It mimics human learning,” she says, noting the algorithm improves in the same way a child learns by experience not to touch a hot stove.
Current 4G (LTE) networks can achieve at best 20-millisecond (millionths-of-a-second) latency, the time it takes to route a data packet. With 6G, wireless researchers hope to achieve “the dream of one-millisecond latency and even less,” says Erol-Kantarci, a speed that might enable applications such as wireless control of cars, mobile virtual reality gaming or 3D meetings.
“Right now, we are determining which parts of a 6G network can best be optimized using AI,” she says, noting that the work builds on her broader research into creating smart networks, including smart electricity grids.
At the School of Electrical Engineering and Computer Science, the wireless engineer runs her own smart human network, the Networked Systems and Communications Research (NETCORE) Lab. “I consider myself very fortunate that I’m in such a hotbed of technology in Ottawa,” she says. Through NETCORE, Erol-Kantarci has established research partnerships with world-leading wireless companies Ericsson, Nokia, Ciena and BlackBerry, among others.
With 14 graduate and post-graduate researchers, NETCORE has become a talent pipeline — sought after by both students and industry, meaning that Erol-Kantarci herself has to juggle daily what wireless engineers call “massive connectivity.”
“It’s a busy environment, but I really like it,” she says. “I really enjoy working with the students, especially seeing the light in their eyes when they come up with some new solutions to the most challenging problems of wireless networks.” Or perhaps, in the not-too-distant future, she’ll be seeing the light of discovery in their holograms’ eyes.