At this year's Brooklyn 5G Summit, NYU professor Ted Rappaport gave a presentation about initial research for what could become 6G sometime around 2030 to 2035. Now, you can read the details in "Wireless Communications and Applications Above 100 GHz: Opportunities and Challenges for 6G and Beyond." Published by IEEE, this paper is available for free download.

In his presentation, Rappaport noted that 5G took 15 years to reach initial deployment and he assumes the same for 6G. Why go beyond 5G? The paper explains that faster wireless speeds will be needed to keep pace with ever-increasing computing power and will create new opportunities. By 2036, we could have $1000 computers that have the computation power of the human brain. Although wireless networks based on terahertz signals still won’t be fast enough to keep up with that power, it will get us closer. Perhaps 7G will get there.

Research at NYU Wireless, a program started by Rappaport, is looking at frequencies above 100 GHz with channel data rates of 100 Gbps. Testing is possible in the U.S., given that the FCC has released 21.2 GHz of spectrum above 95 GHz.

What will it take to get to 6G? A lot of research, both in the electrical and biological domains. In the electrical domain, terahertz signals will present new problems but have the potential for new applications such as sensing that aren’t possible with 5G signals. For example, the ability to "see" around corners and making it possible to sense positions of people in rooms. Even so, much work will be needed to characterize terahertz channels because at such short wavelengths, the roughness of building materials, for example, becomes a factor in absorbing or reflecting the signals. Imagine some building materials acting like the walls of anechoic chambers. Figure 1 shows Rappaport's slide depicting signal losses in common building materials.

Figure 1. Signal losses created by common building materials could result in future materials developed to minimize signal losses at terahertz frequencies. Source: NYU Wireless

We have typically been led to believe that signal attenuation always gets greater as frequencies increase. That's not necessarily the case. Take attenuation in rain. Research has shown that attenuation levels off at about 100 GHz, as seen in this image presented by Rappaport at the Brooklyn 5G Summit (Figure 2).

Figure 2. Rain attenuation levels off at 100 GHz. As frequencies increase, loss from rain does not. Source: NYU Wireless

Another electrical issue that arises at terahertz frequencies that's also wavelength related has to do with antennas and electronics. That is, the antennas get so small that their electronics is the limiting factor in size and thus, electronics may not be integrated into antennas as they are at today's 5G frequencies of 28 GHz and 39 GHz. In fact, heating of components will be more of a problem as engineers try to decrease the size of amplifiers and other components.

Power amplifiers for terahertz signals will have greater noise issues than those operating below 100 GHz. Compensation for these issues may come from the antennas through a concept called spatially oversampled antennas. Spatially oversampled antennas produce "cones of silence" defined as an antenna array's cone-shaped region of support (ROS). The design goal will be to move noise and other undesirable factors into an area outside of the usable field. Such circuits could be based on sigma-delta ADCs and DACs where feedback loops are used to improve resolution. The paper explains this concept in detail.

A discussion of mmWave and terahertz signals would not be complete without mentioning health issues and the need for further study. On the biological front, the paper's authors say that "heating is believed to be the only primary cancer risk" but much work is needed to "understand the biological and molecular impact of THz radiation on human health, even though THz is three orders of magnitude lower in frequency that ionizing radiation," that being X-ray radiation.

Martin Rowe covers test and measurement for EDN and EE Times.