The Millimeter Wireless Frontier

Published in IEEE Spectrum Magazine, March 2017

There is an evergreen quality to the growth of technology.  As existing areas get overworked, new frontiers open up and technology rushes in to occupy the new territory.  There is an example of such a frontier today in wireless communications.  The 5G wireless initiative proposes serving many more users with much higher transmission speeds.  But with the existing cellular bands tightly packed, where does all the needed additional capacity come from?

In contrast with the spectrum management view of scarce capacity, communication theorists see wireless capacity as virtually unlimited.  Capacity can be increased indefinitely by going to ever smaller cells and higher frequencies that offer more bandwidth, while more efficiency can be achieved with advanced signal processing and new spectrum sharing policies.  Among these alternatives the greatest immediate impact would be achieved by moving to the higher frequencies in the millimeter wavelengths -- the region of 30-300 GHz, where bandwidth is available and plentiful.

However, there is a reason for thinking of millimeter wave wireless as a frontier.  Today that band is largely uninhabited and inhospitable, as these wavelengths lie in a most difficult propagation regime.  Free space attenuation increases with frequency so usable path lengths are short, roughly 100-200m. Such distances would be in keeping with the smaller cell sizes envisioned in 5G, but there are numerous other impediments.  Building materials and objects, including people, block the signal.  Rain and foliage further attenuate, and diffraction – that can bend longer wavelengths around objects – is far less effective.  Even surfaces that might offer specular reflection at longer wavelengths appear rougher to millimeter waves and instead diffuse the signal.

So there may be gold in that frontier, but it is going to be very difficult to mine.  Nevertheless, you never know until you try.  I’m reminded of Marconi’s successful transatlantic transmission when physicists insisted that the signal would fly off into space.  Recently a team at NYU has been experimenting with millimeter wave transmission within the urban canyons of New York City.  Like the physicists of yesteryear, I would have said that this would never work.  But their data show otherwise.  There is a surprising amount of coverage, in spite of the shadowing of buildings, pedestrian and vehicular traffic, and the general chaos typical of dense cities.  Granted, there are a number of holes in the coverage, but initial results are encouraging.

When I expressed some surprise at these findings, someone pointed out to me that millimeter wave propagation in its line-of-sight dependence might be likened to that of visible light, and that the nighttime world isn’t as dark as might be expected.  Taking this analogy to heart, I prowled my house on a dark night, illumination coming only from a weak light at the end of a long corridor.  I discovered dim light in unexpected places.  I wondered: how did the light get here?  Even so, nearby rooms might be caves of darkness.  All the while, I was conscious of the strong WiFi signal that followed me everywhere I went.

So this millimeter wave frontier is going to be a difficult one, but we engineers are good at this kind of challenge, and we’re not without tools.  For one thing, at these small wavelengths, we can build postage-stamp sized phased array antennas, and high speed electronics allows us to use advanced processing such as MIMO and Massive MIMO that have been pioneered at lower frequencies.

All this sophisticated technology so we will be able to view 3D video while walking down a busy city street.  Or hopefully some other use.