Multiband, co-located RF installations are placing increasing demands on antennas and RF front ends. There are solutions, but they are not easy to apply
Every wireless link or function has an antenna, of course, and the multiplicity of diverse antennas at a single installation or even buried in a product makes it especially difficult for RF engineers trying to keep them electrically separated. Look at the antennas on the Boeing 787 Dreamliner and you’ll quickly see one challenge, Figure 1.
Some of these antennas are receive only, such as the one for GPS, but a substantial number of them are for duplex links. This means that co-channel and co-location interference are significant issues, with antenna siting and per-channel filtering being very critical as well. If you doubt how much effort is devoted to this problem, just flip quickly through the pages of various microwave-focused magazines or web sites and look at the ads. About one-third of them are for RF-bandpass filters (and another third are for connectors and cable assemblies), which tells you how important these functions are in implementing the physical designs.
It’s not just aircraft with antenna multiplicity. All but the most basic of today’s automobiles also have a variety of antennas for GPS, connectivity, Wi-Fi, radar, emergency calling, and even broadcast AM/FM radio (though not always for the latter, as the EMI in HEVs and EVs is so intense that some manufacturers are leaving the AM radio out, and apparently only a few people are missing it).
The antenna solution for many cars is the “shark fin” multiband antenna and housing, Figure 2, sometimes supplemented by antennas embedded in the window glass. The automotive situation will become even more challenging as Vehicle-to-Vehicle (V2V) and Vehicle-to-Everything (V2X) links become are adopted and proliferate.
A single, multiband antenna can partially solve the “antenna farm” problem, of course. It is not a new idea at all and has been used since the earliest days of wireless. For example, resonant “traps” are used to allow a single antenna to function across widely separated bands. But the new interferers are much closer to each other in the spectrum, and such traps are not adequate.
What can be done? There is no simple, single answer; there rarely are in complicated engineering situations. Careful EM modeling is needed to assess the placement options and tradeoffs for these antennas and their radios, including the impact on RF front ends of all these possible interfering sources.
There’s also considerable work being done to deal with the multiantenna, multiband operation along several fronts. Microstrip PC-board antennas allow modeling and construction of antennas with integral elements that are not realizable as discrete-element designs. Ceramic-substrate antennas are also offering new options, and MEMS switches and passives are being used for on-chip antenna tuning. Companies such as Fractus Antennas S.L and Antenova Ltd, among others, have developed innovative, tiny, multiband antennas as well.
Nonetheless, the fundamental problem of interference from a nearby source to a co-located receiver remains, and is especially difficult when it is a full-duplex situation. The good news is that there’s progress in that area as well. For example, a company called Kumu Networks claims to have a viable, sophisticated technique for suppressing the interference that a transmitter presents to a co-located receiver, even if there is zero guard band between them. In simple terms, they have apparently devised a way to dynamically perform active-RF cancellation, similar to what an audio noise-cancelling system does (see “Self-Interference Cancellation for Co-Located TDD Radios Sharing the Same Band”), except that it is much more difficult in the dynamic RF regime.
Although antennas often get less attention that active RF devices on both transmit and receive sides, they are as important as any other component. As RF links become increasingly ubiquitous and multiband, antenna R&D is also making strides. But it’s also another manifestation of the “no good deed goes unpunished” phenomena: as more solutions are developed to solve the problems, the demands on the system subsequently increase and even more solutions need to be developed. The technology innovation/demand is never-ending.
Have you had to deal with extreme multiband or co-located RF situations? How did you solve the problem? Or did you not solve it?