- Written by Steve Nichols and originally published in Inflight magazine
Interest in airborne connectivity keeps on growing. With the choice of L-band, Ku or the newer Ka, there is no shortage of companies trying to tempt you, whether you operate a commercial airline or a business jet.
But whatever the system, they all have one thing in common – they need an antenna to work. That simple-looking white blob on top of the fuselage or tail fin is often ignored, but is actually a marvel of technology. So what are the problems involved with designing something that has to work in all weathers, mustn’t affect the aircraft’s aerodynamics too much, and yet still has to provide an adequate signal to and from a satellite that could be nearly 36,000km away?
The basics of an antenna are quite simple. It captures radio waves that pass over it, converting them into electrical energy. The most basic antenna is a dipole – two pieces of metal that are cut to an exact half wavelength of the radio waves you want to capture.
The size is all important. When the antenna size matches the wavelength, as illustrated above, we say that the antenna is resonant. This means it is more efficient at converting the radio wave energy. Think of it like pushing a child on a swing – time your pushes to match the natural timing of the swing and it will go much higher. Make your antenna resonate with the incoming radio wave and it will transfer the energy more efficiently.
That’s the basics, but satcom antennas are so much more than that. The higher frequencies mean that different constructions techniques must be used and you also need antenna gain – lots of it – to capture the weak signal coming from the satellite.
There are a number of ways of doing this – at the microwave frequencies used by satellites we can arrange for a whole host of tiny antenna elements to be wired up together, or phased, so that the signal is boosted. Or we can use a dish to focus the radio waves, a bit like using a parabolic mirror to focus the sun’s rays.
Think of antennas in phase as being like a tug-of-war team, all pulling together at the same time. By altering the electrical phasing between antenna elements you can also shape the beam or provide more gain.
Or by altering the electrical phasing between a myriad of different antenna elements on a panel you can distort the antenna’s receive and transmission patterns. This means that you can now electrically “steer” the beam.
At microwave frequencies we can also use a parabolic reflector to focus the beam, or use an array of “horns” or apertures phased together to channel and focus the signals. An array looks like a precision-milled metal potato waffle.
And still it gets more complex. The problem of pointing or steering the beam is compounded when you consider the different latitudes an aircraft has to fly. At high latitudes the satellite is down on the horizon, while over the equator it is overhead. Whatever you design has to be able to cope with both extremes.
There is an added problem where the antenna may interfere with neighbouring satellites due to a wide beam width at some elevations. The solution can be to fit a second antenna or make it dual-panel.
Each satcom system works on a different frequency and so a different wavelength. So an antenna designed for one band won’t work on another and vice versa. For example, Inmarsat’s SwiftBroadband (SBB) works on the L-band, with a frequency of about 1.6 GHz. This equates to a wavelength of about 19 cm. But others, such as Panasonic/AeroMobile, Row 44, Viasat Yonder, and Gogo’s (intercontinental) Ku-band systems work on about 10.9 – 14.5 GHz, with a wavelength of about 2-2.5 cm. Step up to Inmarsat’s upcoming Ka-band GX Aviation system or Viasat’s Exede running at about 20–30 GHz and the wavelength is less than 1.5 centimetres.
The odd-ball in all of this is Iridium. Its satellites are not geostationary, but in low-earth orbit, only about 485 miles (780 km) in altitude. And the frequencies they use are also L-band. As a result the antennas can be simpler, smaller, lighter and lower in gain than antennas for geostationary satellites, but the data throughput (bandwidth) is far less than Ku/Ka or even Inmarsat SBB.
Iridium antennas don’t require complex mechanical or electronic steering and can even be like a small hockey puck allowing them to be used on aircraft as small as a Cessna 172.
With Inmarsat’s SwiftBroadband there are really three classes of antenna – low gain antennas (LGA or Class 15 SBB), intermediate gain antennas (IGA or Class 7 SBB) and high gain antennas (HGA or Class 6 SBB).
The low-gain antenna, such as the Cobham LGA-3000 or Honeywell Aspire 200, is the type you normally find on smaller bizjets. Like a shark’s fin, these are very light and enable SwiftBroadband speeds of up to 200kbps.
Intermediate gain antennas, such as the Cobham IGA-5006 or Esterline CMA-2200SB, might be found on larger bizjets and smaller airlines and enable up to 332 kbps.
The very largest high-gain antennas, such as the Honeywell eNFusion AMT-700 and the Esterline CMC CMA-2102, are really for wide-body airliners or large business jets and enable the full 432kbps service.
L-band has been around for a while, so has all the development work been completed? Far from it says John Broughton, Honeywell Director, Marketing and Product Management (formerly of EMS Aviation).
“No, despite all the excitement and buzz surrounding Ka-band we are still developing L-band antennas,” said John.
“We have a new class 7 intermediate gain antenna for SwiftBroadband that we are bringing to market later this year, probably at NBAA, and we are also working on a new antenna for Inmarsat’s upcoming enhanced SB200 service.”
The new enhanced SB200 service (as yet unnamed) is being introduced after Inmarsat completes its RAN 4 upgrade later this year. By slightly improving the antenna spec for SB200 it will be possible to offer service down to a satellite elevation of five degrees rather than the existing SB200’s 20 degrees. Aviation safety services via the service should also then be possible using a smaller aircraft antenna.
“Despite the interest in Ka, L-band is going to continue to be there, supporting safety services and smaller platforms like business aircraft and helicopters,” said John.
But what about Ku-band? These antennas tend to be steerable types, often with an accuracy requirement of about one degree.
Last year, Panasonic Avionics spent several million dollars to build an in-house Ku-band antenna manufacturing facility at its headquarters in Lake Forest, California. Previously its antennas had been manufactured by Starling and EMS.
Todd Hill, Director GCS Product Management and Capacity, Panasonic Avionics, said that the antenna is the most critical part of the satcomms system.
“Repairing or servicing an antenna on an aircraft is expensive and difficult – you have to take it out of service and get on top of the fuselage. We really wanted to take more control over the design, manufacture and quality control of the antenna.
“So we built our antenna facility in Lake Forest, Southern California, and they are now cranking out antennas faster than ever – and they are excellent quality as well,” Todd said.
Todd added that Panasonic is now producing more than 50 Ku-band antennas a month at the plant. The design is an active array rather than a dish and feed horn. That is, it is composed of a myriad number of tiny antennas all working together (phased), the whole panel being mechanically steered.
“At a radio frequency level maintaining the tolerances needed to make this work effectively is a challenge, one that requires extreme precision,” Todd said.
“The design has to be repeatable to keep the performance levels up. Our single panel also includes everything, including the amplifiers. Having the amplifiers so close to the panel means that losses can be kept to a minimum and they can run at a lower power and be more reliable.”
Todd said that electrically-steered antennas (using the “magic” phasing method, rather than being mechanical steered) are finding favour with the military, but they are expensive and don’t work well at high latitudes, where the satellite is low on the horizon.
“You can electrically bend the beams to point at the satellite, but you can only bend them so far,” he said. “They are being used on L-band, but there are design limitations at Ku and above. We are working with some suppliers, but I think it may be a couple of years before we are able to show one of these actually working on an aircraft.”
Obviously, another problem is that one antenna can’t work on different bands, although some have developed dual-band antennas.
Antennas for Inmarsat’s upcoming Ka-band GX Aviation service bring added design considerations. The first is that the Ka microwave signals are easily degraded by anything that gets in their way – including water, leaves and even the material of the radome.
The mere presence of the radome itself can affect the radiation properties of the enclosed antenna, such as loss and distortion of the radiation pattern. This means that special attention is being paid to the materials being used to make the Ka-band radomes. With Ku-band, a honeycomb construction is possible with a layer of laminates on the outside. But with Ka you need a foam core – and the material must also be transparent to the higher frequencies.
Todd Hill at Panasonic also said that the radome drag with a Ku-antenna can be worse on a narrow-bodied airliner than a wide body one, but it does depend upon the aircraft type itself and the location of the antenna on the aircraft.
The aiming accuracy with a Ku or Ka-band antenna, whether mechanically or electrically-steered, has to be more accurate in view of the narrower antenna beam, compared with an L-band antenna.
Honeywell has the exclusive contract to develop antennas and terminals for Inmarsat’s GX Aviation Ka-band service on business aircraft and air transport. It is currently developing two different antenna types – one is fuselage mounted for the air transport market and the other tail mounted for high T-tail business aircraft.
“The tail antenna is a steered dish.” said John Broughton. “Similar in size to the kind of Ku-band antennas flying on Gulfstream and Bombardier aircraft right now. The fuselage-mounted antenna (on display for the first time at AIX Hamburg) fits under the same radome as the original antenna for the Connexion by Boeing service – both will be mechanically steered.
The antenna designs have gone through the performance development review stage and are on target to enter service, fully certified, in March 2015.
“You want something that is small, light, inexpensive, and aerodynamic, but at the end of the day they have to work, and they have to work really well,” said John.
“They also have to operate down to really low elevations and right now, with today’s technology, mechanically-steered antennas are the only way to get there.
“The first antennas for Connexion by Boeing were phased arrays [electrically steered], but were hugely expensive, tremendously heavy and with low-elevation angle performance that didn’t deliver what people needed.
“Its interesting that when Boeing launched its commercial product they went away from phased arrays and back to a mechanically-steered solution,” John said.
But technology is still advancing.
Honeywell recently signed a contract with QEST to supply Ka-band antenna apertures for integration into the fuselage-mounted airborne antenna system for GX. Developed from QEST’s bi-directional Ku-band antennas, the company’s Ka-band aperture designs feature greater performance and are based on the company’s horn array technology, but the individual apertures (think potato waffle again) are tiny compared with Ku.
Inmarsat also recently announced that it had signed an agreement with Kymeta to develop a revolutionary satellite antenna, enabling business jets to access high-speed broadband connectivity globally through Ka. The Kymeta Aero Antenna will be developed as a lightweight, flat-panel.
“Our technology for flat-panel, beam-forming antennas will enable a number of new markets and a new generation of customers to benefit from lower cost, high-speed satellite internet connectivity anywhere in the world,” said Vern Fotheringham, CEO of Kymeta.
Kymeta’s GX-capable antenna will offer the promise of being significantly lighter and smaller than previous satellite antennas. Metamaterials can magically make an antenna behave as if it were much larger than it really is.
The antenna will electronically steer the antenna beam without requiring power-consuming phase shifters or mechanically-moving parts, reducing the overall cost and power consumption.
It would be easy to think that Inmarsat’s Global Xpress was only the only Ka-band player in town, but that isn’t so. Viasat’s Exede service also uses Ka and is due to be rolled out on United and Jet Blue.
Its VR-12 Ka-band satellite antenna system also recently passed its DO-160G testing. The antenna is designed for use on aircraft such as Gulfstream, King Air, Pilatus, and C-130.
“We designed the VR-12 Ka system to be a form and fit interchange with its Ku-band counterpart,” said Paul Baca, VP and GM, ViaSat Global Mobile Broadband Systems. Commercial airliners will use a different Viasat Ka antenna.
In terms of fitting a satcom antenna, many are available as factory fit on both commercial aircraft and business jets.
But with bizjet retrofit, Brian Wilson, Director of Avionics at Banyan Air Service in Fort Launderdale, said: “On SwiftBroadband business aircraft installations by far the most common installation is the Low Gain Antenna for the SB200 service. We have installed this system on aircraft ranging from a Citation 500 to a Boeing 727.
“For aircraft that have a centre-mounted engine like a Falcon 50, 900, you are very limited to fuselage antennas due to ice build-up that could fly into the centre engine. Most new fuel-efficient aircraft have two engines so this problem is eliminated.
Another consideration is that possible interference from other aircraft systems means that the antenna has to be mounted a certain distance away and/or filtering may need to be installed at the antenna.
But what about the time needed to fit an antenna? “A low gain antenna (LGA) should take about two-three days, while an intermediate gain antenna (IGA) should take about four-seven days,” said Brian. “But fitting a Ku-band antenna to the fuselage could take longer, and modifying the top of a vertical fin to accommodate a bullet radome could take three-four weeks.
“My feeling is that once Inmarsat introduces the enhanced SB200 service and lowers the coverage from the current 20 degrees line of sight to five-degree coverage (which is the same as HGA) it will dominate the bizjet market,” Brian said.
Brian added that in addition to the Inmarsat SB200 systems they are also fitting a lot of Aircell Air-to-Ground (ATG) systems and antennas for use in America/Canada.
So spare a thought for that innocuous white blob that hides the satcom antenna on top of an aircraft. It is hiding something that really is a technological marvel.