Top 5 Flat Panel Antenna Myths

I’m looking forward to joining the panel for the GVF Webinar on Transformational Antennas on 13 August! An enormous amount of investment in LEO constellations presumes the availability of low-cost, electronically-steered antennas, but it is important to be realistic about what they can and can’t do.

To set the stage, here are my Top Five Myths:

1. They are optional. Nope – we need them! If we don’t have low-cost electronically-steered antennas, LEO communications constellations just won’t succeed.
2. Flat panel antennas will make parabolic antennas obsolete. Highly unlikely! If a parabolic antenna suits the application (fixed terminal on a GEO satellite), it will always have better performance and lower cost than a flat panel antenna. Why? For a given size, small parabolic antennas are extraordinarily cheap, and they have close to theoretically-perfect performance. If you rarely need to repoint the antenna, the cost of 10-20 minutes of extra installation time for manual pointing is much less than the higher cost to make an electronically-steerable antenna. For LEO service, though, the beam must follow the satellite and switch quickly to the next one, so electronic steering is very important.
3. Flat panel antennas are smaller because new technology is more efficient. False. Flat panel and parabolic antennas are just two cases of aperture antennas. In all aperture antennas, the gain, beamwidth, and sidelobes are determined by the size and area facing the satellite, the illumination function, and feeder losses. Parabolic antennas are close to theoretical optimum for losses and illumination (taper). Array antennas often have significant feeder losses, which reduce transmit gain and receive G/T. Plus if they are pointed more than 10-30 degrees from their zenith, gains quickly degrade. Therefore, flat-panel antennas generally have lower efficiency, so for the same performance, must be larger, not smaller.
4. Flat panel antennas give horizon-horizon coverage with no moving parts. Unfortunately, this is impossible. To close the link budget and prevent interference, the antenna must present a minimum width and height projected along the direction of the satellite. As a flat antenna’s beam is steered electronically away from its zenith (i.e. the “scan angle”) the effective area towards the satellite is reduced. For example, at 45 degrees signals reduce by 50% (-3 dB), and beyond that, performance degrades rapidly. That’s why many LEO constellations are planned to have huge numbers of satellites – that way there are always one or two close to the zenith, no matter where you are, so the antenna does not have to scan very far.
5. Flat panel antennas can prevent adjacent-satellite interference by controlling sidelobes. Sorry, the laws of physics say no. For any small aperture antenna (under 1.5m at Ku-band) the main beam is wide enough to interfere with satellites 2-5 degrees away. The width of the main beam is determined solely by the cross section of the antenna across the beam direction, and the illumination function (taper). You can’t narrow the main beam width by tweaking the array’s illumination function. (You can, however, affect farther-out sidelobes and nulls, which may be beneficial in other ways.) And reducing interference by spectrum spreading requires significantly more satellite bandwidth. Ideally, LEO constellations are planned and coordinated with each other so that satellites on the same frequency and pol never come too close to each other. This can, in theory, allow “undersized” user terminals to coexist with each other and GEO satellites too.

---