The electromagnetic spectrum is the highway over which wireless operates, with multiple lanes capable of carrying traffic at different speeds. Higher frequencies – and thus shorter wavelengths – are able to move more information per unit of time.
Millimeter or extremely high frequency (EHF) waves, occupy the relatively unused portion of the electromagnetic spectrum between 30 GHz and 300 GHz, which offers greater throughput and thus higher overall capacity than the increasingly crowded WiFi bands under 6 GHz.
Historically, millimeter-wave technology has been expensive and difficult to deploy, which has limited it to niche applications like radio astronomy, microwave remote sensing and terrestrial fixed communications.
More recently, however, interest has increased significantly as those two obstacles have been largely overcome. Millimeter wave has evolved into a cost-effective option for meeting the ongoing network capacity challenge faced by enterprises. We expect prices to continue to decline and price/performance to continue to improve, making millimeter-wave solutions a first option for organizations everywhere.
A wide range of outdoor and indoor applications are poised to benefit from this set of technologies.
The traditional fixed point-to-point (P2P) and point to multi-point (P2MP) microwave communications are all right at home in millimeter-wave solutions. Millimeter-wave products offer little additional complexity over microwave-based solutions, and licensing, where required, is almost always handled by the equipment vendor or dealer involved in the sale and installation. Parabolic antennas used in millimeter-wave applications can already be seen in campus and metro-area building-to-building bridges, backhaul and interconnect, links to ISPs, and in ad-hoc applications like telemetry and surveillance.
Wireless PAN and LAN
The IEEE wireless personal area network standard 802.15.3c, the 802.11ad Wi-Fi standard and upcoming 802.11ay standard all specify the 60 GHz band. Given the limited object penetration, range and directionality of millimeter-wave signals, applications here are usually restricted to in-room and open-office settings where line-of-sight can be assured. We expect that general access applications using 802.11ad will become common; chipsets are now making their way to market.
We also expect an increasing variety of outdoor and campus solutions based on these technologies. Many existing microwave and millimeter-wave solutions will evolve over time to inexpensive P2P, P2MP and mesh solutions based on .11ad components, opening the door to even wider deployment opportunities.
Similarly, we expect millimeter-wave components to find their way into a variety of IoT solutions, given that so many of these will be clustered and thus amenable to short-range and mesh approaches.
Despite the potential for multi-gigabit throughput, 802.11ad has seen essentially no real-world uptake to this point. This is due largely to the success of 802.11ac in addressing pent-up and growing demand for WLAN capacity, but also suspicion regarding the behavior and utility of the 60-GHz bands – wariness that is eerily similar to that expressed regarding the 5-GHz bands when 802.11a was initially introduced in 1999.
Given time for users to travel up the experience curve and the attraction of low prices and excellent price/performance, this situation is likely to improve over the next few years.
As 5G is intended in large measure as a replacement for all other wireless WANs and even, where economically feasible, wired broadband services, significant new backhaul capacity will be required. Since 5G will often apply the “small cells” model of denser base-station deployments, millimeter-wave backhaul makes sense even with its inherent range limitations, and even in the unlicensed bands, thanks to narrow beams, directional antennas and the applicability of mesh techniques enabling rapid, cost-effective deployment of backhaul capacity.
More controversial is the application of millimeter waves for 5G mobile subscriber access. Some experiments have indicated that this should be a valuable option in locales with densely packed base stations and appropriate beamsteering capabilities. It’s not clear, however, if the industry will adopt millimeter-wave mobile access on a large scale, and appropriately equipped subscriber devices would be required and, obviously, are not currently on the market.
Millimeter-wave radar is already in use in some automotive applications, and mobile connectivity at these frequencies might be useful in inter-vehicle links for network-based vehicles of the future.
Millimeter waves are also useful in many location and tracking applications, both short-range for desktop pointing-device implementations, and outdoors for more macro-scale situations. Uncompressed HDMI video is also often cited as a key application, although given the high-performance and low-cost video compression standards, algorithms and implementations available today, such is likely of decreasing importance.
Finally, some research has been done into intra-and inter-equipment-rack communications using millimeter waves. While we view this as somewhat exotic, consider that the cost of such links could be significantly less than that of fiber with no compromise in performance. Similar chip-to-chip links on circuit boards have also been proposed.
Advances in millimeter-wave technology
Millimeter waves have historically been very difficult to generate, and building the required oscillators and other components have traditionally involved the application of exotic and expensive semiconductor processes, most notably gallium-arsenide (GaAs) – a key reason why lower frequencies have been utilized to a greater degree.
In recent years, however, we’ve learned how to design and manufacture components capable of operating at millimeter-wave frequencies using more cost-effective processes, like silicon-germanium (SiGe), and especially the same complementary metal-oxide semiconductor (CMOS) processes commonly used in processors, memory chips and most other components related to computation and communications today. CMOS is inexpensive, with high yields, excellent across-the-board performance, and easy integration into radios and related devices like OEM modules.
In short, we’re now able to utilize millimeter-wave-based communications reliably and cost-effectively, but it’s still not quite the smooth sailing of the more established radio bands.
There’s clearly demand for what millimeter waves can offer, but there are additional constraints that need to be considered.
Propagation characteristics – Millimeter waves tend to be highly directional, propagating in narrow beams and usually blocked by solid objects like the interior and exterior walls of buildings. For this reason, most applications of millimeter waves are line-of-sight (LoS). The effective range for a given application is a function of how clear the desired path between endpoints is, the transmit power applied and the type and configuration of the antennas used. It’s certainly possible to use directional antennas and even techniques like MIMO (which depends upon multipath), beamforming and even beamsteering via active (“phased”) antenna arrays to enhance throughput and extend range. Antennas, even directional parabolic dishes, are usually fairly small, given the tiny wavelengths involved.
Limitations resulting from the fundamental physical behavior of waves at millimeter-wave frequencies, however, still apply. Signal attenuation at many millimeter-wave frequencies isn’t a major concern, but the 60 GHz band is unlicensed because 60 GHz is the oxygen-absorption frequency – that point in the electromagnetic spectrum where the oxygen in the air attenuates radio waves.
This can be compensated for by using multi-node deployments with shorter distances between nodes, with the significant benefit of improved overall capacity via frequency reuse. Given narrow beams and directional antennas, interfering with other connections on the same frequencies nearby can usually be avoided.
Other millimeter-wave bands also exhibit varying degrees of atmospheric attenuation, but many – most notably 20-50, 70-90, and 120-160 GHz are only slightly impacted. In short, given a good match between a given band and the right applications (see below), the limitations of millimeter waves can usually be overcome and reduced to the point of insignificance.
It should also be noted that the narrow beams, limited range and related behaviors of millimeter waves actually serve to enhance security and integrity. Regardless, encryption must be applied on all wireless links, and most commercial millimeter-wave products support encryption.
Regulation and licensing – Both licensed and unlicensed bands are defined at millimeter-wave frequencies. Among the most common commercially available licensed millimeter-wave bands are 27-31, 38, 71-76, 81-86, and 92-95 GHz, and we can expect the spectrum-auction process to be increasingly applied to these bands.
Solution topologies – The millimeter-wave bands have been used primarily for fixed applications, where both endpoints of a given link are stationary – the P2P and P2MP applications. Going forward, we expect to see much greater application of mesh techniques, which offer the greatest degree of flexibility and coverage possible in any network configuration.
Millimeter-wave wireless, once an exotic and expensive option for outer-space and terrestrial fixed communications, is now a cost-effective and in many cases inexpensive option for meeting the ongoing network-capacity challenge.
For enterprise IT, there are implications for upgrades to both in-building and campus wiring and Ethernet switches going forward in order to get the most out of an investment in millimeter-wave communications.
But the potential for much higher capacity, as well as even-higher throughput when needed, makes the millimeter wave fast lane the place to be.