Love Story: President Goldsmith Explores Past, Present and Future of Wireless

When were the seeds of wireless technology planted? What have we learned? And most importantly, what might the wireless future hold? Stony Brook President Andrea Goldsmith addressed these questions and more on April 17 in her Inauguration Week lecture, "How I Fell in Love with Wireless."

Goldsmith shared with an enthusiastic audience in the Charles B. Wang Center Theater her early beginnings as an engineering math student at UC Berkeley in the 1980s and wanting to do something practical.

A mentor suggested communications because it was a mathematical side of engineering, and could also have an interesting practical impact. "That was a lesson in how a mentor can have a big impact on your life, because if he had not suggested that, I never would have had the incredible opportunities I've had to work in this amazing field."

Graduating in 1986, Goldsmith recounted her entry into a field that barely existed at the time. She joined a tech company in Silicon Valley, but the defense industry had collapsed and there was no work in commercial wireless communication because cell phones had just come out. Despite those challenges, she was able to open a door at Maxim Technologies to a soon-to-be-burgeoning field, having no idea of the contributions she would make in the years to come.

"I fell in love with wireless communication," she said. "The notion that you could communicate over wireless channels anywhere in the world and imagining what that might be just captivated me."

Though WiFi was still years away, she got to work with some of the most visionary wireless experts at the time who thought about problems and approached them differently than she did.

The discussion touched on technology challenges like the spectrum deficit in cellular communications, particularly the insufficient bandwidth for data-intensive activities, the challenges of signal power reduction and the importance of ultra-reliability and low latency for applications like telemedicine and autonomous driving.

Noting that 5G came with a lot of promises, Goldsmith recalled a favorite quote by Jonathan Swift: 'Promises and pie crust are made to be broken.'

"That's especially true in wireless standards, because those promises drive all the work that goes into the next generation of standards," she said. "Right now, there are a lot of people working on the 6G standard, and in 2030 we'll have that."

To that end, Goldsmith spoke of a wireless future that includes smart homes and spaces, autonomous cars, smart cities, body-area networks and robot teams, as well as the problem of resource allocation, a complex problem without a good mathematical model.

She offered a brief history of AI and wireless communication, highlighting the evolution of AI from the 1950s to the 2010s, emphasizing the shift from computational to deep learning-based AI.

The conversation also covered the impact of AI on wireless systems and the potential of AI in 6G for dynamic optimization and resource allocation.

"6G calls itself 'AI-native,' which means it will use AI to do a kind of optimization in the cloud to figure out the user patterns," she said. "And if you think about using AI for this cloud optimization, it should also apply to every network like vehicle networks and sensor networks and drone networks. AI is really going to play a big role in wireless networking optimization."

Goldsmith recounted her initial resistance to embracing AI, describing herself as a skeptic when it comes to technology hype.

"Everybody was working on it, I didn't see how it was going to be useful for wireless communication, and I'm not a 'bandwagon' kind of researcher," she said.

Eventually she was swayed by the research she was doing, which convinced her that AI would play an important role in the next generation of wireless design. Chemical communications showed that a deep learning algorithm could beat theory.

"One of the things that that taught me at a higher scale was that AI works really well when we don't have good mathematical models or we have complexity constraints," she said. "This is where AI has demonstrated that it can play a big role, and that's why I believe that this AI-based cloud optimization is what we're going to see in this next generation of networks."

Goldsmith concluded by comparing Stony Brook's place on the leading edge of quantum computing to those at UCLA in 1969 who helped create the Internet and extend it across the country. Stony Brook has established a quantum Internet running across Long Island, with plans to expand it to upstate New York with the $300 million investment from New York State to drive the quantum revolution, before moving across the country.

"So you'll see that the quantum internet started here at Stony Brook will eventually connect every quantum device throughout the country and around the world," she said. "This is an exciting time for wireless communication, maybe the most exciting time ever because of the applications it will enable."

- Robert Emproto