Siddharth Ramachandran, a BU College of Engineering Distinguished Professor of Engineering, has shown corkscrew-shaped beams of light could share up to 50 times more data than current networks.

Photonics

Could Twisting Beams of Light Make the Internet Faster—and Greener?

BU engineer Siddharth Ramachandran is developing a technology that could cut the power needed to transmit data—and maybe one day improve brain imaging

You may not think that hours of scrolling on TikTok or searching on Google impacts the environment the way that driving a gas-guzzling car does, but everything you do online sucks up energy and resources—and comes with a climate cost. With a majority of the planet now online, the 150 to 1,000 grams of carbon dioxide generated by every hour of video calling or streaming quickly adds up. Some have estimated that four hours of watching Netflix can have a similar carbon impact to driving a car one mile. And artificial intelligence is only intensifying the power drain, requiring vast resources to fuel its data-heavy needs. Microsoft reportedly struck a deal earlier this year to fire up a defunct nuclear power station just to propel its cloud computing and AI systems.

With more and more data centers popping up across the world—and more and more online content being shared and watched every day—could we make our digital world greener?

“This is an extremely complex problem that requires a massive convergence in multiple planes and axes—scientists, engineers, planners; everything needs to come together,” says Boston University electrical and computer engineer Siddharth Ramachandran. He’s working on a technology that could potentially help make the internet faster and more efficient—and cut its power consumption.

Ramachandran is an expert in optics and photonics—the use and manipulation of light. It’s light that carries the data zipping down fiber optic cables to and from your home or office. For more than a decade, Ramachandran, a BU College of Engineering Distinguished Professor of Engineering, has been developing a way of using twisted beams of light to more reliably and efficiently transmit data. He’s found the corkscrew-shaped beams, a form of structured light, could share up to 50 times more data than current networks.

“We’ve demonstrated the ability to use optical fibers that will allow you to send much higher bandwidth information,” says Ramachandran. The technology could be ready to apply in as little as five years.

Twist the Light Fantastic

Turn on a flashlight, laser pointer, or car headlamp and the light pours out in a straight line. Sunlight does the same thing. It’s what light does—keeps on trucking until something gets in its way or bounces it off in a new direction. The structured light beams Ramachandran works with are the exception. By giving photons a twist—“selectively changing the direction in which different parts of a light beam point”—he gives the beam what’s known as optical angular momentum.

“It travels in a helical path, exactly like the wind structure of a tornado,” says Ramachandran, who was this year named an American Association for the Advancement of Science (AAAS) Fellow and editor-in-chief of the leading journal Optics Express. “The physics of its stability is a lot like that for a spinning top or a bicycle wheel—if we don’t spin it fast enough it falls, but it becomes more and more stable the faster it rotates.”

And that stability becomes a big deal when trying to cut down on interference with communications—depress a current optical fiber carrying light, for example, and the light distorts, and so does your signal.

“Anything that has very high rotational momentum, called angular momentum, is by itself stable to several other perturbations in the system,” says Ramachandran. “And, at the end of the day, communication is about how to reliably send information from point A to point B without it getting corrupted.”

When he first heard about structured light as a member of the technical staff at Bell Laboratories, Ramachandran was told it was an interesting scientific quirk with few, if any, practical applications. But it sparked an idea and he began to study, then test, ways of using it to transmit information. A few years after joining BU in 2010, Ramachandran and one of his students, Nenad Bozinovic (ENG’08,’13), published a paper in Science that showed the spinning beams could double, perhaps even quadruple, fiber optic cable capacity.

It prompted a rush of media interviews and has since been cited by other scholars more than 3,000 times. “That woke the community up to the fact that optical fibers can transmit exotic states of light,” he says.

In 2023, Ramachandran and his team—including Zelin Ma (ENG’23)—pushed the load even higher, demonstrating in another Science paper that they could outperform current optical fibers’ data handling capacity by 25 to 50 times.

From More Efficient Internet to Better Brain Imaging

Before your internet is brought to you by tornadoes of light, Ramachandran has a couple of barriers to overcome. He’s already making progress on the first one: ensuring that the energy required to create the structured light doesn’t wipe out the savings made in transmission efficiency. “The biggest challenge is that these states of light are harder to create, harder to maintain,” he says, “and anything that’s harder consumes energy to create it.”

And because generating the spinning light is complicated, getting to the point where industry can “stamp out devices that the Verizon engineer can slap onto the end of a wire” will take some work. “That doesn’t worry me too much,” he says. “I am relatively confident that if this becomes a deployed technology, especially by one of the big telecom suppliers, then the cost and complexity of such add-on devices will come down very, very quickly.”

Ramachandran says that because running any new optical fiber is a major (and costly) logistical enterprise—land and roads need to be disrupted, communications interrupted—it tends to happen in 10, even 20-year cycles. While that may slow full implementation of the twisted light technology in what he calls the “long-haul transmission fibers,” other advances could come sooner.

“There are regeneration centers spaced between 25 to 100 kilometers apart, where you take the signal that is distorted and recoup it, then build it up,” says Ramachandran, who also holds appointments with BU’s physics department and centers for neurophotonics, photonics, and systems neuroscience. “The innovation will first go in these regeneration centers, in the amplifiers. One of the things we’re working on is rather than doing very long-haul transmission with these twisted beams, we are using them to amplify. The timescale for that is maybe 5 to 10 years.”

He’s also exploring other uses for the technology, from quantum computing to brain imaging. One in-progress study suggests that structured light may allow scientists to see deeper into the brain than ever before. Ramachandran is collaborating with biomedical engineers and neuroscientists at BU and beyond to test that neuroimaging potential.

“It would be absolutely awesome,” he says, “if our technologies could help solve one of the biggest problems, which is to understand how the brain works.”

Notable funders of Ramachandran’s research include the Defense Advanced Research Projects Agency, Air Force Office of Scientific Research, and Office of Naval Research, as well as the National Science Foundation and National Institutes of Health.