There’s something about antennas that fascinates me. Perhaps it’s their functional simplicity: they do one thing, they’re tangible (whether formed from carefully shaped and dimensioned conductive material, as microstrip embodiments, or perhaps as a metallized ceramic resonator), and best if all, they require no initialization or software to function. Academically, their role is to act as bidirectional transducers between the wired and wireless worlds, via Maxwell’s basic equations of electromagnetic principles.
Antennas have a very wide range in size, largely (but not entirely) dependent on frequency and transmit-power level. They can be tiny, as those embedded in a smart phone, or huge, such as the VLF antennas spread over thousands of acres, pumping out tens of megawatts in the 10 kHz-and-below region (yes, that’s kilohertz) to reach submarines when submerged (the dB/meter attenuation through sea water is severe and increases with frequency). There are also developments in the use of metamaterials which offer new, previously “impossible”, performance possibilities.
Even though there are a near-infinite number of basic and advanced antenna designs in widespread use, there’s a continuous need for new ones, see Figure 1 . There are also highly specialized ones such as those used on spacecraft which unfold after launch; some CubeSats use antennas made of the springy metal strips used for standard tape measures, a simple and cost-effective solution.
It won’t come to this (hopefully), as designers are creating compact multiband antennas which are sufficiently efficient and can meet both immediate needs as well as emerging ones such as support of some 5G bands. (Image source: Harman International (Samsung), via Front Seat Driver)
Still, the need for new application- and mission-specific antennas continues, and it’s not just due to the increased use of parts of the spectrum that were almost empty until now. Recently, the US Army Combat Capabilities Development Command, C5ISR Center issued a two-page RFI (Request for Information) “W56KGU-19-R-MULTIBAND – Multiband Antennas” seeking ideas for two new multiband-antenna designs. (Note that RFI here does not stand for “radio frequency interference” although, of course, an antenna could be used for that!)
For two-port antenna #1, they would like coverage on Port 1 from 600 MHz to 1000 MHz; 1400 MHz to 2025 MHz; and 2100 MHz to 2700 MHz, with a 300-W power rating. For Port #2, the band is just 2400 MHz to 6000 MHz and a 50 W rating. For two-port antenna #2, the ranges at Port 1 are the same, while Port #2 should handle 2400 MHz to 6000 MHz; no power rating is provided). Adding to the challenge, both antennas must meet military specifications for physical ruggedness, temperature tolerance, and other harsh conditions.
For many years, the process of antenna design was a largely intuitive skill combing theory, experience, and instinct, perhaps with a little “magic” added to the mix. Start with a whip or dipole (plain or folded), then add, subtract, enhance, and try; then cut/fit/trim and repeat. Of course, you also have to size the antenna for the transmit power levels (if any) and adjust the elements to compensate for fringing and parasitic-like effects which can degrade and shift performance from expected values suggested by theory and experience values. If it’s a multiband antenna or multiple co-located single-band antennas, the gap between the theory and reality becomes wider.
In the past few years, however, antenna modeling and simulation tools have advanced significantly and become very sophisticated. They can take Maxwell’s equations and use them to create very good models of proposed designs and also take real-world considerations into account as well. This has dramatically increased the rate at which new designs can be assessed with a high level of confidence, since testing physical antennas in an anechoic chamber is costly and time-consuming.
The EM solvers can analyze, but they can’t (yet) figure out radically new configurations – that’s still the province of experienced antenna designers who have a sense of what is worth trying and the inevitable tradeoffs. That imagination limitation may erode in the next few years due to AI, but there’s no way to know what will happen or how fast it will happen.
If you look at RF-centric publications, at least one-third and often one-half of the contributed articles are from research and university authors showing new antenna configurations for both established and emerging applications across various regions of the spectrum; amateur-radio publications such as QST also feature many articles showing new designs as well as new implementations of old ones.
In my view, there are several reasons for this, in addition to the obvious importance of the subject. First, the inherent nature of antennas means that new configurations are easily imaginable, especially when compared to, say, new structures for a solid-state device. Second, it’s relatively inexpensive to build a test model (whether discrete metal or PCB-microstrip design, compared to doing that solid-state device. Third, the cycle time for devising, stimulating, and fabricating the new antenna configuration is short enough to be a good university project. Finally, doing a comparison between simulated and actual model performance is very feasible.
Have you had to devise a new antenna structure for a >1 GHz application? What were the RF and physical issues that most surprised you?
Related content ,/p>