UW-Madison researchers are fast becoming leaders in technology to enable an increasingly interconnected world.
There are more devices on earth capable of connecting to the Internet than there are people living on the planet.
Phones in people’s pockets, wearable technologies on their wrists, smart devices in their homes, and self-driving cars all tap into wireless networks. Those networks exchange more than 3.7 exabytes (or 1018 bytes; one byte is roughly equal to a single character) of information every month, which is equivalent to 80 percent of the data volume of the entire Internet, or enough digital material to fill up five and a half billion compact discs.
That staggering volume of information puts tremendous strain on wireless networks—and the modern world’s invisible infrastructure is long overdue for an update. “The need for advances in wireless communication and mobile systems is stronger than ever before, especially as users continue to expect anytime, anywhere access through their personal devices,” says Suman Banerjee, a professor of computer sciences and electrical and computer engineering and head of the Wisconsin Wireless Networking Systems Laboratory.
But slower-than-molasses data-rates aren’t just an annoyance for 90,000 fans trying to post photos to social media at a sporting event, or even people simply out running errands; an overloaded network could prevent some important emerging technologies from ever seeing the light of day. “There are a lot of new innovations such as smart cities, autonomous vehicles, and the Internet of Things, which are resting on the premise of fast access to cloud computing and the cloud being everywhere,” says Akbar Sayeed, a professor of electrical and computer engineering who leads the Wireless Communications and Sensing Laboratory.
Sayeed is among many UW-Madison researchers who are developing transformational technologies that rely on wireless to tap into the cloud—the data stored and accessed anywhere through a wireless connection.
For example, students working in the Internet of Things Lab at UW-Madison design smart devices such as wearable monitors that can contact emergency medical professionals if a patient’s vital signs vary from healthy values. In 2015, a system of wireless sensors that ensures people use proper form during strength-training exercises moved beyond the lab and into the UW-Oshkosh football team’s weight rooms for field-testing under the auspices of WeightUp, a startup company co-founded by UW-Madison electrical and computer engineering graduate student Pete Chulick.
Smart devices also are coming online in developing nations. Electrical & Computer Engineering Professor Giri Venkataramanan’s research group is working to develop small-scale electricity distribution systems called microgrids that empower rural communities to take ownership of their own energy generation. The circuit boards that manage microgrids come with wireless connectivity, meaning that people can text-message their homes to activate cooling systems or monitor energy consumption from afar. Thanks to a partnership with the startup company NovoMoto, founded by engineering mechanics PhD students Aaron Olson and Mehrdad Arjmand, several prototype microgrids are already up and running in the Democratic Republic of the Congo.
Right now, third-generation (3G) wireless networks cover most industrialized nations, and fourth-generation (4G) systems are becoming increasingly common. Those 4G networks are considerably faster, but engineers are already looking ahead to bring about the next generation: 5G. When the fifth generation of wireless does come online, the speediest 4G systems will seem glacially slow by comparison. “We can imagine that a 5G wireless Internet connection will be as fast as a wired connection. Cellular connections will be as fast as ethernet cables and a lot of devices will be connected to the 5G network,” says Electrical and Computer Engineering Assistant Professor Xinyu Zhang.
According to standards set by the International Telecommunications Union (ITU), 5G wireless will achieve gigabit-per-second data rates for multiple users in the same area. In other words, no more foot-tapping while you wait for a page to open in your phone’s Internet browser: Imagine sending or receiving the complete works of William Shakespeare—20 times—in a single second. That’s a gigabit. In more modern terms, a 5G connection would enable you to download an entire episode of a 30-minute high-definition television show in less than 30 seconds.
Reaching the fifth generation won’t be easy. Each of the preceding generations built on progress that didn’t require fundamental changes to the wireless infrastructure.
“This is a really exciting time. 5G will be a clean slate because it is so different. The basic science and the engineering are not well understood, and neither is the technology. Here at UW-Madison we’re pushing forward on all fronts.” —Akbar Sayeed
The first generation of wireless cell phones, for example, made calls using analog modulations of radio signals on 1980s-era phones like the Motorola DynaMAX8000 that stood as tall as bricks, and weighed nearly twice as much.
Finland launched the world’s first second-generation (2G) network in 1991. The first to convert voice calls to digital information, it paved the way for modern data messaging. In fact, just one year later, in 1992, a Canadian engineer sent the world’s first text message, which read, “Merry Christmas.” Now, Americans send more than six billion text messages every single day.
Although some industrialized nations, like the United States, Singapore and Switzerland, have announced plans to eliminate second-generation networks by 2017, many people in the developing world still use 2G. However, because the systems operate far below the 200 kilobits-per-second data rates that define 3G, browsing the Internet on a 2G connection can be painfully slow.
Third-generation wireless (3G) delivered the necessary performance to surf the web from mobile devices, but the initial roll-out was slow after debuting in Japan in 2001. When smartphones hit the market, however, more people started demanding mobile broadband and now 3G covers 84 percent of the world’s population.
The ITU established standards for 4G systems in 2008, but eased back on those initial requirements in 2010. Even though the Long Term Evolution (LTE) and WiMAX networks coming online in the United States and Asia often fall short of 100 megabits per second mobile data rates that define true 4G, these systems use different frequency bands and deliver substantial performance improvements. “In the first four generations, people relied a lot on improving communication hardware and signal processing algorithms, but that aspect has reached an upper limit,” says Zhang.
He predicts that 5G will use more disruptive technologies—for example, millimeter wave.
And without those disruptive technologies, there simply won’t be enough bandwidth to go around. The Federal Communications Commission has strict rules in place to divvy up radio frequencies, which is a good thing because crossing frequencies among different applications would be disastrous. Pilots communicating with air traffic control towers use different frequencies of radio waves than pop music stations, for example.
Waves on the electromagnetic spectrum can be described by the distances between their crests. These wavelengths can range from smaller than atoms, like gamma rays, or miles long, like time signals that set radio clocks. Currently, mobile wireless receives a few frequency bands in the neighborhood of 1.7 and 1.9 gigahertz, which means those waves range from about 2.5 to 6 inches—about the length and width of a dollar bill.
Waves on the electromagnetic spectrum can be described by the distances between their crests. These wavelengths can range from smaller than atoms, like gamma rays, or miles long, like time signals that set radio clocks.
Because all electromagnetic waves travel at the speed of light, huge numbers of radio waves vibrate past a given point in one single second, and the frequency of those waves increases as the wavelength gets smaller. Tiny ultra-high-energy gamma rays cycle more than 10 billion-billion times per second (a whopping 10 exahertz), whereas ultra-low frequency waves that vibrate a mere 50 times per second span more than three miles in length (1 cycle per second equals one Hertz).
AM radio broadcasts use waves about 450 yards long. They oscillate 10 million times per second, or 1 megahertz. Over-the-air television broadcasts occupy frequencies around 800 megahertz, corresponding to a wavelength of about the distance traveled during one average walking step.
Currently, mobile wireless receives a few frequency bands in the neighborhood of 1.7 and 1.9 gigahertz, which means those waves range from about 2.5 to 6 inches—about the length and width of a dollar bill.
But even space for electromagnetic waves is limited, and some bands are filling up fast. “Existing wireless networks operating below 6 gigahertz simply do not have enough spectrum to sustain the current growth and deliver the expected data rates,” says Sayeed.
Fundamental theories dictate that the speed of a wireless network is directly proportional to the amount of available bandwidth—the distance between the lowest and highest frequencies in a given range. Bandwidth does not come cheap. A typical 4G cellular tower uses around 100 megahertz of spectrum to cover everybody in the area, and that 100 megahertz usually costs billions of dollars.
In July 2016, the U.S. Federal Communications Commission made waves in the wireless community by freeing up a whopping 11 gigahertz (one gigahertz is 1,000 megahertz) worth of spectrum specifically for 5G and mobile broadband. Two days later, the National Science Foundation announced a seven-year, $400 million initiative to make 5G a reality.
But taking advantage of those newly available regions of spectrum won’t be as easy as switching stations on a car radio. The majority fall in very high frequency ranges—above 24 GHz—which is in the millimeter wave range of the spectrum. Unlike the longer, lower frequency waves used now, millimeter waves can’t pass through many obstacles such as walls or human bodies.
That requirement for an unobstructed path is among myriad challenges for 5G wireless. But ongoing research at UW-Madison is turning the Badger State into something of a wireless research hotspot. “There’s a critical mass coming together in Wisconsin, even though we’re far away from Silicon Valley or Telecom Valley,” says Sayeed.
Wisconsin engineers are not only developing new hardware and software, but also setting up testbeds to help bridge the gap between laboratory prototypes and market-ready products.
In fact, Zhang’s laboratory was one of the first places to ever send streaming video over 60 GHz millimeter waves on a platform called WiMi. Using that testbed, Zhang’s group figured out a potential solution to obstructions in millimeter wave travel: Reflect the signal beam off a solid nearby object and steer around the obstacle.
WiMi is but one of many tools on campus through which researchers can tinker with new hardware and software schemes for 5G wireless. Another testbed led by Sayeed is designed to devise strategies for maintaining connections with multiple moving users using state-of-the-art multi-input/multi-output antennas that use multiple beams for communications. This testbed is based on a patented wireless technology called CAP-MIMO, pioneered by Sayeed’s group, that is also being developed for potential commercialization in emerging millimeter-wave 5G applications.
And newly renovated lab spaces on the third floor of Engineering Hall, updated thanks to a generous gift from ECE alumnus Peter Schneider (BSEE ’61, MSEE ’63 PhDEE ’66), are already starting to buzz with activity. “These new labs (in combination with the remodeling of the Plexus Collaboratory and the Qualcomm design labs) are a fantastic asset for the department’s research in mobile systems,” says Electrical and Computer Engineering Professor Parmesh Ramanathan, who researches advanced computational algorithms capable of processing millimeter wave signals into useful information.
The research isn’t only confined to labs on campus; citizens in Madison have been field-testing advanced wireless systems ever since 2010, when two Madison Metro buses began connecting passengers to high-speed internet, thanks to a project called WiRover, led by Banerjee. Now Banerjee is setting the wheels in motion on a new initiative, named WiNest, that promises to turn the entire city of Madison into a testbed for advanced wireless and mobile systems.
Sayeed and Zhang also are heading a research coordination network in the area of millimeter-wave wireless to foster collaboration among industry developers, academic scientists, policymakers, and federal stakeholders as these normally disparate groups all start exploring the bold new frontiers of 5G wireless. “This is a really exciting time,” says Sayeed. “5G will be a clean slate because it is so different. The basic science and the engineering are not well understood, and neither is the technology. Here at UW-Madison we’re pushing forward on all fronts.”