|Last Updated: Fri Mar 29 10:48:39 UTC 2013|
Electronically Steered Arrays
A Maturing Technology
by Dr Carlo Kopp, MIEEE, MAIAA, PEng
First published in Australian Aviation
The advent of production Active Electronically Steered Array (AESA) radar antennas represents one of the most important, if not the most important development in radar technology since the 1940s. With unprecedented reliability, superior performance and typically of the order of a one thousandfold improvement in beamsteering speeds, this technology will transform many aspects of air combat and strike operations.
The idea of an electronically steered antenna is not new. Early warning radars used for the detection of ballistic missile attacks have exploited this technology widely since the 1960s. The US Navy's SPY-1 Aegis radar, developed during the 1970s to defend carrier battle groups against saturation attacks by Soviet cruise missiles, is perhaps the best known electronically steered antenna design in operational use.
The B-1B Bone has flown since the 1980s with an AN/APQ-164 radar, fitted with an electronically steered array. The B-1A Batwing also exploits this technology in its AN/APQ-181 multimode attack radar. Both of these radars can be used for terrain following flight, as well as surface attack modes. The Soviet MiG-31 Foxhound also carries this technology in its SBI-16 Zaslon air intercept radar.
The important and recent development in electronically steered antennas is the solid state X-band (centimetric wavelength) AESA, built using Gallium Arsenide chips. This technology is now being retrofitted to in-service fighters and will be a standard production component in the F-22A and F-35 series fighters, and most likely a downstream production/upgrade item for late model F-16C, F/A-18E/F, Typhoon and Rafale fighters.
Active Electronically Steered Arrays - A Primer
To best understand the importance of the AESA it is useful to explore the limitations of conventional mechanically steered antennas, and the first generation of passive Electronically Steered Arrays (ESA).
The basic purpose of any microwave radar antenna on a fighter (or bomber) is to focus transmitted microwave power into a narrow beam, or receive reflected microwave power from targets (or terrain), again within a narrowly focussed beamwidth. Targets are found by steering the antenna repetitively through a programmed pattern, to search a volume of sky or a surface footprint. The antenna transmits the energy, which travels out to the target, reflects and is then received by the antenna. For an antenna to be useful it must therefore not only be capable of launching and receiving microwave power, but must be steerable precisely and preferably very quickly.
The one bit of good news in antenna physics is the reciprocity theorem which says that the radiation pattern of an antenna when transmitting power is of the same shape as its reception pattern.
In an ideal world the antenna produces an absolutely sharp beam - the radiation pattern is for instance conical, and all energy transmitted is within that cone, and all energy detected by the antenna is also within that cone. In the real world this of course does not happen, the radiation pattern leaks outside the main lobe of the beam, creating what are termed sidelobes. A good measure of antenna quality is how small the sidelobes are eg ten times (10 deciBels or dB), one hundred times (20 dB), or one thousand times (30 dB) below the mainlobe in magnitude.
The first generation of microwave fighter radar antennas were mechnically steered concave reflectors, colloquially termed dish antennas. This is the basic technology the RAAF still operates in the F-111's AN/APQ-169 and -171 radars. These antennas have several drawbacks - they are expensive to fabricate to high accuracy, tend to have fairly large sidelobes, and also have a frequently large radar signature when illuminated by a hostile radar - as all concave reflecting cavities do.
By the 1970s the state of the art shifted to mechanically steered planar array or slotted array antennas, an example being the AN/APG-65/73 radar in the RAAF's F/A-18A. Planar arrays achieve their focussing effect not by reflection as concave antennas do, but rather by manipulating the individual time delays into a very large number of very simple slot antennas, arranged in a planar array panel. By using a cleverly designed and oft complex network of microwave waveguides on the rear surface of the array, a designer could produce the desired fixed beam shape and do so with much smaller sidelobes compared to a concave reflecting antenna. As the antenna is a flat plate, it tends to act like a flat panel reflector to impinging transmissions from hostile radars and thus has a lower radar signature than a concave antenna.
While the US focussed on planar arrays, the Europeans and Soviets deployed a number of Cassegrainian reflector antennas on fighter radars, as these performed better than concave reflectors but were cheaper to design and fabricate than planar arrays - the design and manufacture of the complex feed networks on the rear face of any array antenna is still considered to be somewhat a black art.
Planar arrays provided important gains in beam quality but due to the need to mechanically point them they remained slow to steer and also suffered the same reliability problems as concave antennas. The complex mechanical gimbal arrangement and servomotors used to drive such antennas suffer from wearout, and the cyclic mechanical loads on the antenna proper can also induce material fatigue failures over time.
Airborne radar designers covetously eyed the electronically steered antenna technology used on ground based radars and by the 1980s this technology found its way into airborne radars, some examples noted earlier. An electronically steered antenna of this ilk is designed with an individually electronically controlled device behind each antenna element, which can manipulate the time delay or phase of the microwave signal passing through it. With a computer controlling each element in unison, the beam direction and its shape could be digitally controlled, within a matter of milliseconds or tens of milliseconds.
The first generation of such antennas used the signal phase as the controlling parameter, typically using ferrite core devices for this purpose. Therefore such antennas were known as phased arrays. A typical design could resemble a conventional planar array, but with a layer of digitally controlled ferrite core phase modules inserted between the antenna array elements and the microwave feed network on the back of the antenna. As the antenna contained only what engineers term passive components, these antennas are also known as passive phased arrays or passive electronically steered arrays. The Russian SBI-16 Zaslon, Phazotron Zhuk Ph and NIIP N-011M are good examples, as are the AN/APQ-164 and -181 mentioned earlier.
This generation of Electronically Steered Arrays permitted unprecedented beam seering agility compared to mechanically steered antennas, and very large reductions in antenna radar signature if well designed. Beamsteering agility in turn permitted the important capability of interleaving radar modes. An ESA can timeshare multiple and diverse modes, a good example being shared concurrent operation performing both terrain following and surface mapping for weapon delivery.
However, the antenna did nothing to enhance either the reliability or the efficiency of the radar. The high power Travelling Wave Tube in the transmitter remained, causing its traditional share of reliability woes, while the complex interconnections required to connect the digital control signals to the wirewound ferrite cores was an additional burden. Since the ferrite cores introduced signal losses in both the receiving and transmitting directions, these antennas were less sensitive than their mechanically steered predecessors, and required more powerful microwave tubes to drive them.
The US DoD recognised the need for a better antenna technology more than two decades ago. A new technology, using the phased array concept but with a miniature transmitter and receiver in each antenna element, was seen to be the answer to the limitations of existing technologies. Known as active phased arrays or AESAs, these antennas became the holy grail in the radar community - for reasons yet to become fully apparent.
The enabling technology for AESAs is the Gallium Arsenide Microwave Monolithic Integrated Circuit (GaAs MMIC) or microwave circuit on a single chip. GaAs MMICs would permit the low cost mass production of AESAs, with high reliability and repeatability.
Gallium Arsenide is however a finicky material to make chips from and it took almost two decades for the fabrication technology to move from expensive botique manufacture to industrial strength mass production. Today this technology is being put into cellphones, broadcast satellite receivers and TV sets. The author recalls a development project in 1984 where he has not permitted to use a GaAs low noise transistor in a $100k piece of high speed communications equipment - too expensive, Carlo, find a cheaper way to do this!. A decade ago the GaAs component market was dominated by military buys, which today comprise only around 2% of the total market volume.
The problems in producing a digitally controlled solid state AESA were evident very early - cost, density and power handling would be critical. All of these factors have contributed to the relatively late deployment of the technology in operational aircraft.
The basic building block of any AESA is the Transmit Receive Module or TR Module. It is a self contained package making up one AESA antenna element, and contains a low noise receiver, power amplifier, and digitally controlled phase/delay and gain elements. Digital control of the module transmit/receive gain and timing permits the design of an antenna with not only beam steering agility, but also extremely low sidelobes in comparison with passive ESA and mechanically steered antennas.
Two other important benefits are derived from this design approach. The first is an very important improvement in antenna noise behaviour, since the TR module's low noise receiver is within the antenna itself. Typically this yields between a two and fourfold reduction in receiver thermal noise, in turn contributing to improved radar sensitivity and thus detection range, all else being equal.
The second important benefit is a result of the transmitter power stages being distributed across hundreds or over a thousand TR modules, rather than being concentrated into a single transmitter tube. As a result, failure of up to 10% of the TR modules in an AESA will not cause the loss of the antenna function, but merely degrade its performance. From a reliability and support perspective, this graceful degradation effect is invaluable. A radar which has lost several TR modules can continue to be operated until scheduled downtime is organised to swap the antenna. Other beneifts also accrue - a classical design with a high power tube must carry the transmitter power to the antenna through a pressurised waveguide, and power will be lost in the antenna. Transmitter tubes require highly stressed high voltage high power supplies, which also tend to be unreliable. Since each TR module only handles several Watts of power, fed from a low voltage supply (several Volts rather than kiloVolts), it can be designed for much lower electrical stress levels.
As a benchmark, typical conventional fighter radars have Mean Times Between Failure (MTBF) of around 60 to 300 hours - AESA radars push the MTBF into the 1,000 hours or better class. Rather than several repairs annually to the radar, the AESA will see the radar needing repair only once every several years of operation. If we assume an annual flying rate of around 200 hours, on average the AESA needs to be repaired once every five years! From a support costs perspective, this means much reduced cost of ownership for fighter fleet operators who transition to the technology.
An AESA becomes a combined transmitter, low noise receiver and beamsteering package, providing high beamsteering agility, very low radar signature when illuminated and extremely low sidelobes, all digitally controlled. With digital control of TR module gain, power management which is vital for reduced or low probability of intercept (RPI, LPI) operation becomes relatively easy to do. Beamsteering agility also facilitates reduced or low probability of intercept scan patterns.
In many respects an AESA is a fighter radar designer's dream device, since it not only vastly improves performance and functional capabilities, but does so with improved reliability and complete digital control of antenna/transmitter functions. Over the life of an AESA radar, progressive refinements in many aspects of antenna behaviour can be incorporated through incremental software upgrades. Software programmable AESAs at this time largely implement digital equivalents of established antenna beam shapes, scan patterns and sidelobe behaviours. Over time with proper intellectual effort, further improvements are possible.
Are there any drawbacks to the AESA? Two issues are of key importance with this technology.
The first item of interest is power dissipation. Due to the behaviour of microwave transistor amplifiers, the power efficiency of a TR module transmitter is typically less than 45%. As a result, an AESA will dissipate a lot of heat which must be extracted to prevent the transmitter chips becoming molten pools of Gallium Arsenide - reliability of GaAs MMIC chips improves the cooler they are run. Traditional air cooling used in most established avionic hardware is ill suited to the high packaging density of an AESA, as a result of which modern AESAs are liquid cooled. US designs employ a polyalphaolefin (PAO) coolant similar to a synthetic hydraulic fluid. A typical liquid cooling system will use pumps to drive the coolant through channels in the antenna, and then route it to a heat exchanger. That might be an air cooled core (radiator style) or an immersed heat exchanger in a fuel tank - with a second liquid cooling loop to dump heat from the fuel tank. In comparison with a conventional air cooled fighter radar, the AESA will be more reliable but will require more electrical power and more cooling, and typically can produce much higher transmit power if needed for greater target detection range performance (increasing transmitted power has the drawback of increasing the footprint over which a hostile ESM or RWR can detect the radar).
Another issue of concern with AESAs is the mass production cost of the TR modules. With a fighter radar requiring typically between 1,000 and 1,800 modules, the cost of the AESA skyrockets unless the modules cost hundreds of dollars each. With early module builds yielding unit costs of around USD 2,000 the cost penalty of using an AESA over a conventional design was prohibitive. The good news in this respect is that the ongoing trend has been downward, in a large part as the production engineering of the modules and MMIC chips has improved. Having an enormous commercial market for similar MMIC chips has yielded important benefits.
The longer term technology trends for AESAs are clear - a progressive cost reduction as volumes increase and production matures, with concurrent refinements in digital antenna control techniques improving the capabilities of the antennas.
At this time the US are leading the pack by a large margin in AESA technology, with the EU and Israelis trailing. The Russians remain in the passive AESA domain but this will change as commercially available GaAs MMICs proliferate. The Russians have a robust track record in passive ESA design and the only obstacle to an AESA equipped Su-30M is the availability of suitable chips in volume.
Current AESA Programs
A number of AESA programs are currently under way, some as new build radars for new fighter designs, some as retrofits to existing aircraft.
The first generation of AESAs to field will be the L-band (decimetric) radars used in the Israeli Phalcon derivatives, and importantly the RAAF's Wedgetail MESA radar which is expected to be used in the new USAF MC2A (Multi-sensor Command Control Aircraft) E-3 AWACS, E-8 JSTARS, RC-135 Rivet Joint replacement. The lower L-band operating frequency of these radars permits the use of older transistor technology, giving this class of AESA about 5 years of market lead against the X-band fighter AESAs.
The E-8 JSTARS is a candidate for an X-band AESA replacement of its existing passive ESA radar, used for high resolution mapping and surface target detection. It is likely that the planned JSTARS AESA will be used on the MC2A - the MC2A variant to be used to replace the JSTARS/AWACS combo will be a 767 airframe carrying both radars (this is a potential growth path for the Wedgetail, by the addition of a JSTARS/MC2A derivative radar under the forward fuselage).
US sources indicate that the RQ-4 Global Hawk is likely to see an AESA upgrade later in the decade, to provide increased range and radar optimisations for accurate ground target tracking, and airborne target tracking
In the fighter domain the first AESA to field is the
AN/APG-63(V)2 on the F-15C. This is a major upgrade of the original
AN/APG-63 radar using a 1500 element AESA. While only 18 F-15Cs were
originally to be retrofitted, the gains in reliability and thus reduced
support costs are likely to see this AESA migrate onto the USAF's 200+
strong F-15E fleet, as well as further F-15Cs. The Korean F-15K is
expected to carry this AESA, making it the most capable fighter in the
The next AESA to field will be the Northrop-Grumman AN/APG-77 radar on the USAF's F-22A. This 1,500 element AESA will remain the highest performing fighter radar in the market for the forseeable future. Designed with a very low radar signature antenna, it will provide the F-22 with a greater detection and engagement range than any other fighter in the market. Current planning envisages that the -77 will transition from the current TR module design to a design common to the F-35 JSF radar, with the aim of cost reductions. Recent US disclosures suggest that this will happen around the middle of the decade, as an F-22A Block 5 configuration, with the AESA rework incorporating antenna features for undisclosed advanced ground attack modes. The AN/APG-77 radar is the first of new generation of radars which push the back end signal and data processing functions into the aircraft's central computers, rather than inside a dedicated radar processor box.
The third AESA to field will be the 1,000 element
Northrop-Grumman AN/APG-80, formerly the AN/APG-68 Agile Beam Radar, on
the UAE's new F-16C Block 60 fighter. This radar is a substantial
redesign of the established AN/APG-68 and will utilise COTS VMEbus
derived high performance processing in the radar back end functions.
It is expected to support not only the full range of air-air and strike
modes, but also interleaved terrain following. IOC is likely to be in
the 2003-2004 timeframe.
The fourth AESA to field is likely to be the 1,100 element Raytheon AN/APG-79, formerly AN/APG-73 RUG III, on the USN's new F/A-18E/F fleet. This AESA is a block upgrade to the existing AN/APG-73 series. Whether the whole F/A-18E/F fleet receives the radar has not been disclosed, but given the longer term life cycle cost benefits this could become a long term priority for the USN.
The F-35 JSF, if it proceeds to plan, will field with a 1,200 element AESA radar using a similar architecture to the F-22's AN/APG-77. While little has been disclosed in the way of design details, this radar is likely to resemble a scaled down and less capable F-22 radar, with a strong optimisation for strike roles.
Other players are entering the market. Work continues on the European AMSAR AESA for the Eurofighter Typhoon. Public data suggests a 1,500 element design although this might be optimistic given the size of the Typhoon. An AMSAR derivative is a likely retrofit option for the Rafale.
Israel's Elta has published datasheets on a range of X-band GaAs MMIC chips which would be suitable for an AESA but as yet no disclosures of system level products have been made.
Reports also suggest that the Russian radar houses are working on AESA designs, although details remain very sketchy. Any Russian design would have to make use of imported GaAs MMIC chips as Russia's industry lags severely in this area. A likely outcome is that COTS GaAs MMICs would be adapted for a Russian design, as the export controls on high volume X-band satellite transceiver chips are likely to be unenforcable over the coming decade. A suitcase of GaAs MMIC chips makes for a lot of AESAs.
It is not inconceivable that we may see a robust number of AESA retrofits over the coming decade to established teen series fighter fleets in the West, as the investment in this technology returns a large payback in medium and long term support cost reductions. With large fleets of the F-15C/E and F-16C in USAF and export client service, possibly several hundred aircraft could be retrofit candidates.
What impact does the AESA have for Australia? Clearly any aircraft considered for A6K must have an AESA - anything less is simply an unnecessary drain on fighter support budgets.
In terms of retrofits, the remaining life cycle and increasing strategic irrelevance of the F/A-18A HUG make it a poor candidate for an AN/APG-79 AESA retrofit later in the decade. The F-111 would be an excellent AESA retrofit candidate especially since the support cost reductions and reliability gains over the 1960s AN/APQ-169/171 suite would achieve a major impact in life cycle costs in the 2015-2020 timescale. As recent testimony by LtGen Mueller to a parliamentary committee indicates, the principal issue in F-111 life-of-type will be the support of the remaining 1960s generation aircraft systems in the post 2010 period, given that the RAAF is now retrofitting AMARC F-111F wings to gain significantly more airframe fatigue life.
Current trends are that the AESA will supplant conventional mechanically steered radars in all production fighter applications over the coming decade, and there are good prospects for partial fleet retrofits across the several thousand teen series fighters likely to remain in service over the next 2-3 decades. The only issue with the AESA will be securing near term funding for retrofits, given that the support cost payback may take several years to be seen. Given the tremendous combat capability gains resulting from the AESA, the case for retrofits is very robust if the aircraft is to remain in service for more than a decade.
In summary, the AESA will have a revolutionary impact over the coming decade, and smart players should be now exploring how to best exploit this pivotal technological development.
|Artwork, graphic design, layout and text © 2004 - 2013 Carlo Kopp; Text © 2004 - 2013 Peter Goon; All rights reserved. Recommended browsers. Contact webmaster. Site navigation hints. Current hot topics.|