The light way to 6G

The world of 6G will be one of abundance.

In this future, we live inside an invisible spider web of high-frequency radiations that tie together billions of cellular devices, millions of autonomous vehicles, and trillions of sensors.

Our reality seamlessly merges with the virtual. Our decisions are informed by data, computed by artificial intelligence with zero latency. We converse with holograms, conduct surgeries remotely, and no longer hear the phrase "outside network coverage."

Hopefully, this would also be a future where hunger and poverty are close to being eliminated—one where the wealth of data in the clouds have come to mean an equality of opportunity for those living below it.

The sixth generation of wireless technology—shortened to 6G—sounds a lot like science fiction. However, the systems required to implement this vision are not too far away. Samsung Research estimates that 6G could reach full-scale commercialization around 2030—which is just a decade away.

World over, research into the technology needed to actualize 6G has gathered momentum over the last few years. In 2018, The Academy of Finland announced "6Genesis," an eight-year research program to conceptualize 6G through a joint effort with Nokia. Universities in the UK, US, Japan, South Korea, and Singapore have launched research to meet the insatiable needs of the Internet of Things (IoT), medical devices, sensing, and communications. Industries—large and small—are gearing up to carve out a space for themselves in 6G, even though 5G is only starting to be deployed.

As Daniel Mittleman, professor of engineering at Brown University, puts it, "With the exact parameters and performance of 5G still unknown, it is too early to say what 6G will require." He is a world leader in terahertz technologies—one of the leading candidates for communication beyond 5G—and also the former chair of the The International Society of Infrared, Millimeter, and Terahertz Waves.

It's quite possible that there will be no clear demarcation between 5G and 6G. However, we do know of one requirement that all future wireless communications need to provide: hyperfast data transfer rates.

By the time 6G matures in 2030, the way we communicate with each other could change drastically, incorporating everything from holograms to virtual reality. Most of these systems are very data hungry.

Consider that a realistic 3D hologram of a human face requires 19.1 gigapixels. Updating these points in real time to match gestures and expressions will need a download rate of 1 terabits per second (Tbps)! Creating a virtual replica or "digital twin" of a 1 m x 1 m space needs 0.8 Tbps assuming synchronization every 100 milliseconds (ms). Similarly, truly immersive augmented reality (AR) and virtual reality (VR) require high data rates.

The peak data rate for 5G is expected to be 20 Gbps. As a frame of reference, that is four orders of magnitude higher than the 4G LTE in our current generation of phones.

At the same time, Samsung Research estimates that the global market for AR and VR is expected to reach a total $131.7B USD by 2030, which is an indicator of the data rates that will be required. Clearly, 5G's 20 Gbps will not be enough to satisfy these requirements.

It's not just these futuristic technologies—cell phones and cloud-connected devices are estimated to proliferate until there are over 10 million devices per square kilometer. The global tide of the Internet of Things will create up to a trillion sensors that are all bound by a network. Many of these devices will have built-in "native AI" systems that compute decisions rapidly from available information.

Those that require more powerful AI systems—such as autonomous vehicles—will offload computations rapidly to the web where the results will be processed and returned in real time.

Combined, the global mobile traffic is predicted to reach 5,016 exabytes per month in 2030 (compared to just 7.462 exabytes per month in 2010). Further, the needs of industrial automation, emergency response, and personalized medicine require that these communication channels have high reliability and extremely low latency to enable real-time data processing.

This means that 6G requires high bandwidths—much higher than 5G.

"The relation between the amount of information that can be carried by a channel and its bandwidth is given by Shannon's Capacity theorem," explains Ranjan Singh of the Nanyang Technology University, Singapore, where he explores novel methods for fast data transfer using photonics.

A high-frequency channel has a larger bandwidth and can carry more information than one at a lower frequency. 4G runs in the megahertz or a million cycles per second. 5G will cover a higher frequency range up to a maximum of 75 Ghz.

That leaves us two frequency bands where 6G could operate competitively.

The first is the band of radiation spanning from red to blue of the electromagnetic spectrum, the intimately familiar photons that we identify as visible light. The technology to generate and manipulate them is widely available—a white light LED, for example, costs less than a US dollar. As Marcos Katz, professor at the University of Oulu, Finland, points out, "Light-based wireless communications technology has been researched and developed for a couple of decades."

The challenge however, is that light is easily blocked or absorbed by walls, trees, moving objects, fog, or rain. This severely restricts the range of these systems. "When we talk about optical wireless communications, we refer to two approaches: namely free-space optics (FSO) and visible light communications (VLC). The former refers to a connectivity up to a few kilometers using laser beams, whereas the latter refers to short-range wireless communications typically in indoor environments, with ranges below 10 meters," Katz explains. His work, as part of the 6Gensis consortium, is in this second approach.

The second option is the terahertz band—a band of radiation extending from 0.1 Thz to 30 Thz—where electric and magnetic fields oscillate at the rate of trillions of cycles per second. Because their wavelength is less than a millimeter, they are also called submillimeter waves.

The terahertz band is the dark horse of the electromagnetic family. They are not just less well known than radio or microwaves, there are also fewer devices that can generate and or modulate these waves. With the exception of submillimeter telescopes in astronomy, the terahertz regime is a technological backwater, so much so that this band is often referred to as the "terahertz gap."

"Terahertz systems are much less well developed and, as a result, more expensive," concurs Mittleman. "On the other hand, there are some real reasons that THz is a superior alternative. Issues such as pointing stability, sensitivity to atmospheric turbulence, and eye safety all argue for THz wireless over FSO. Their performance in adverse weather conditions are complementary: THz is better in rain or fog, but FSO is probably better in snow."

It is these reasons that have made terahertz an active area of research in the last few years. In particular, a number of photonic technologies built on silicon such as waveguides, filters, multiplexers and demultiplexers, modulators, antennas, and photodetectors have been developed to meet the technological requirements for this medium.

The photomixer, for example, is a simple photonic device that converts optical radiation to terahertz waves in a single step. Here, two laser beams with different frequencies are made to coincide with each other. The result is a wave with frequency equal to the difference of the base frequencies, which falls into the terahertz regime. Coupled with an antenna, photomixers become an excellent terahertz generator that can produce custom terahertz frequencies.

Another photonic technology that is used to generate terahertz signals is the quantum cascade laser. Unlike a typical laser where the electron falls from a higher energy level to a lower energy level, the electrons in this laser "cascade" through a sequence of energy levels before reaching the ground state. A common quantum cascade laser is made of slices of gallium indium arsenide (GaInAs) and aluminum indium arsenide (AlInAs) that are stacked periodically on an indium phosphide (InP) substrate. The cascading energy states of the electron is a product of this periodicity.

Photonics offer a distinct improvement over electronics in the terahertz band. As Guillaume Ducournau, professor at the University of Lille and a leader of THz wireless communications activity explains, "Photonics is one of the driving technologies for THz/6G. It initiates the highest data rates so far and enables characterizations of advanced devices like active circuits in 6G systems." In 2018, his team demonstrated the first wireless terahertz link using a silicon photonics transmitter. An embedded Ge photodiode generated a 300 Ghz signal at a data rate of 10 Gbps that was received by a commercial GaAs Schottky diode.

Another win for silicon photonics is that it's also compatible with current CMOS fabrication facilities. "This allows semiconductor foundries to mass produce 6G devices without a major overhaul of their facilities," adds Singh. "The key goal is to generate and modulate terahertz signals at the chip level."

This year, Singh's team at NTU, in collaboration with Masayuki Fujita's team at Osaka University in Japan, attained the highest on-chip terahertz wireless data transfer rate ever recorded: 16 Gbps.

They achieved this using a novel concept called a photonic topological insulator (PTI). This is a material that behaves as an insulator on the inside but conducts electrons or photons at the surface. These conduction states are remarkably robust. They survive twists, turns, bends, and deformation.

A topological photonic insulator made by etching alternatively large and small triangular holes on a silicon chip. Credit: Ranjan Singh, NTU

Achieving a photonic topological insulator is remarkably simple. A silicon chip is etched with alternating triangular holes of two different sizes so that they arrange themselves into a hexagonal lattice—much like carbon atoms in graphite. The interface of these triangles with air traps light through total internal reflection, ensuring that all the photons are conducted from one end to the other without any scattering. Using this chip, the teams from NTU and Osaka University transmitted an uncompressed 4K video in real time at the rate of 11 Gbps, while navigating 9 bends. The channel frequency was 335 Ghz, safely in the 6G terahertz band.

Using a more complex encryption technique known as quadrature amplitude modulation (QAM), the collaborating teams from NTU, Osaka, and Lille have recently pushed this up to 75 Gbps at a bandwidth of 25 Ghz. By changing the geometry of the crystal, they are confident of pushing the rates even higher. "The ultimate dream is to get past 100 Gbps using a photonic topological insulator," said Singh.

While terahertz communications is pushing towards new frontiers, some of the old challenges remain. These waves are absorbed by water and oxygen molecules in the air, and lose energy to free space which severely restricts the range of application. To overcome these, high gain antennas are required to send focused directional waves. Innovative data transfer methods are also needed, such as multiple-input-multiple-output (MIMO), which multiples the number of channels at which we receive and send information, increasing the total amount of information between devices.

However, as Katz reminds us, terahertz communications also has a cost barrier. "THz components could be highly expensive as their design is complex and time-consuming, while optical components are well known, studied, and relatively low cost," he says.

Katz's team is pioneering a light-based sensor network designed to collect and exchange information from every part of the human body. "Complementing the IoT, the Internet of the Body (IoB) will keep track of crucial body functions such as heart rate, enzyme levels, and other vital signs. It will safely and securely link implant devices such as pacemakers, gastrointestinal pills, and in-brain devices. Near-infrared signals can easily penetrate up to 5 cm inside the body—enough to reach chest implants—and transmit at 100 Kbps. The hyperfast data rates of 6G are not as much a requirement for the IoB as are safety, reliability, and security.

A light-based Internet of Things (LIoT) is also being explored by Katz and others. The idea is that light signals could connect devices within a room with each other and to the WiFi. To be effective, these signals will need to be directly aimed from the source to the receiver. The advantage is that these devices might be cheap and easily attainable. In aircrafts, visible light communication, also known as LiFi, has been shown to be effective and safe even during takeoff and landing.

In the end, when 6G does come about, it is quite likely that it will use both terahertz and visible light. Katz says, "In the end, optical wireless communications and THz communications could well co-exist in 6G."

Daniel Mittleman agrees. "Right now, if you want to forecast where we'll be in 10 years, there is certainly no clear winner between these two approaches. Ultimately, I think there is a place for both of these technologies—and, in situations where maintaining a link is absolutely crucial, you'd probably want both."

Vineeth Venugopal is a scientist and science writer who loves all things and their stories.