Movatterモバイル変換

Russia's military radarindustryhasadvanced considerably since the end of the Cold War, largely resultingfrom accessto Western technologies in the global market. This has seen significantadvances in basic technology, especially in such key areas such asradarsignal processing, radar data processing, embedded software, GalliumArsenide semiconductors for low noise receivers, and in HEMT (HighElectron Mobility Transistor) transistors used in Active Electronically Steered Arrays (AESA).

This sustained growth in basic technology has been reflected in ongoinggrowth in the capabilities of the various radars deployed in RussianAir Force and export variants of the Sukhoi Flanker fighter.

This analysis will survey the basic radar types available, summarisethe data, and include a cardinal parametric analysis. RepresentativeWesternradar performance will be compared, with a focus on Beyond Visual Rangeair combat regimes of operation.

Pulse Doppler radarsremain theprimary long range sensors used by fighter aircraft for BVR combat.This is for several good reasons. Infrared sensors cannot penetratecloud or other atmospheric propagation impairments as well as X-bandmicrowaves can. Radar, conversely, can penetrate most weatherconditions and impairments from the stratosphere down to the lowestlayers of the troposphere. Effective range is another consideration,as radar performance is limited by the pulsed power-aperture product ofthe design, which in the current state of the art permits X-bandfighter radars to acquire larger targets at distances in excess of 200nautical miles (~400 km). Radars are also capable of rapidly diviningthe velocity, direction, altitude and often identity of targets, whichcan be problematic for passive sensors operating in the optical bands.Radars typically also double up as X-band datalink transmitters forlong range missiles, an important factor in achieving high killprobabilities in BVR combat, where the time of flight of the missilewould otherwise create opportunities for a target to move outside oftheNo Escape Zone (NEZ) of a BVR missile seeker. Forthe forseeable futureradars will remain the primary tool for the acquisition, tracking, andengagement of targets in the BVR air combat game.

The design parameters of most interest to analysts and competing radardesigners in this area are those which determine the ultimate limits onthe detection range of the radar against representative airborne targettypes atlong ranges. These are all contained in the most basic forms of theradar range equation, and the physics of radar performance it describes.

Peak Power (P_peak)[kW]is the maximum pulsed power the radar can emit. It islimited mostly by the transmitter technology employed, and to a lesserextent, the antenna design. In general, the higher the peak poweremitted, to the first order, the better from a range perspective.Peak power is also important in Electronic Warfare terms as itdetermines the burnthough performance of the radar, or the point atwhich the energy reflected by a target is greater than the energyproduced by the target's defensive jamming equipment. This thepoint  where jammingeffectively fails.

Aperture Gain (G) [-]isa measure of the area and efficiency of the antenna employed fortransmission and reception. The bigger the aperture gain in a radar, tothe first order, the better from a range perspective.

Power Aperture Product (PAorPxA) [Wm²,dBWm²,dBW] in its mostcommonly used form is calculated bymultiplyingPeak (or Average) PowerxAntenna Area (or Power [dBW]+ Antenna Gain [dB] in [dBW]). It is a parameter used by designers togauge therelative performance of different radar designs. To the first order,the radar with the higher Power Aperture Product or PA will achievebetter range, detection and jammer burnthrough performance.

Receiver Noise Figure [-]is a measure of the thermal and shot noise effects which are competingin the radar receiver with intended signals to be received. The lowerthe noise figure (or 'noise temperature'), the better. Receiver noisefigures are generally similar for given generations of radartechnology, reflecting the radio frequency transistor types, andantennaconfigurations used.

In practical terms,  to maximise detection range and jammerburnthroughperformance,  the biggest  radars  in terms of power andantenna size win over those with smaller antennas and less power.

Leaving detection and tracking range performance aside for a moment,other radar parameters and attributes are also relevant in a combatenvironment. Unfortunately  these capabilities and parameters areoften not so easy to compare parametrically, and in many situations areless important than the range and burnthrough performance.

Sidelobe Performance[deciBels]of a radar antenna determines how much energy is emitted in directionsother than than intended, and how much energy is detected fromdirections other than intended. Sidelobe performance is important inrejecting ground clutter when pursuing low altitude targets, and inproviding good resistance to jamming. Jammers are often designed toinject false targets into a victim radar via its sidelobes.

Mainlobe Width [degrees ofarc]of a radar antenna determines how narrow the main lobe of the antennaradiation pattern is, or in simpler language, how narrow a'pencil-beam' of microwave energy the radar produces. As the so called'antenna  reciprocity theorem' applies, for a typical antennadesign themainlobe (and sidelobe) parameters are the same for transmitting asthey are forreceiving. For typical fighter radars, mainlobe widths vary between 4°and 2° of arc. For many applications, the narrower the beam thebetter, within limits.

Antenna/Receiver/TransmitterBandwidth is a measure of the radar's potential frequencyagility, or itsability to hop across frequencies to evade detection and jamming.Bandwidth is also important for many modes which require wide bandwidthmodulations in the signal. These include Low Probability of Intercept(LPI) modes, High Power Jam (HPJ) and high speed datalinking (HSDL)modes.

Signal ProcessingPerformance is a measure of how manycomputations the radar signal processor can perform per second on thedigitised raw radar video signals collected by a receiver. Thisparameter is often measured in terms of Fast Fourier Transformoperations executed per second, or where performance is consideredsensitive, in the less revealing measures of MIPS (millions ofinstructions per second). Signal processing performance will impact theradar's ability to sift targets from noise, jamming and low altitudeclutter. A related parameter is the number of receiver channelsemployed. In digital radars these are usually paired. Again, a largernumber of channels is typically better.

DataProcessingPerformance is a measure of how manycomputations the radar signal processor can perform per second on thetarget track data collected by the radar, as well as on computationsassociated with missile guidance and envelope management. Dataprocessing performance will impact the radar's ability to track largenumbers of targets, manage multiple missile engagements, controlmultiple missiles in flight, and perform other functions important tothe managment of the radar's operation.

Beamsteering Agilityis ameasure of how quickly the antenna mainlobe can be pointed in adifferent direction, and/or reshaped for a different operating mode. InMSA(Mechanically Steered Antenna) designs this parameter is typically ofthe order of hundreds of milliseconds. In AESA/ESA (Active /Electronically Steered Antenna) designs this parameter is typically ofthe order of hundreds of microseconds, or a thousand times faster.Beamsteering agility is important in tracking targets, multitasking theradar between diverse operating modes, providing resistance to jamming,and supporting multiple concurrent missile shots.

Angle Tracking Techniqueis the method used by the radar to measure the angular position of atarget within the radar's mainlobe. The favoured technique in recentdecades is monopulse angle tracking, due to its accuracy and resistanceto many jamming techniques. Monopulse angle tracking can use multilobedtechniques, or sequential lobing techniques.

Given the complexity ofradarequipment internally, and from a design perspective, there are manyother parameters which define the overall capabilities of a radar andits performance across the wide range of modes in which modernmultifunction radars are used. However, in long range missileengagements, to the first order of magnitude, the Power ApertureProduct isthe critical parameter.

PowerAperture Product Performance inFlanker Radars Over its long productionhistory a wide range of radar types and subtypes have been fitted tovariants of the Sukhoi Flanker fighter. Whilst some observerslike to generalise the term 'Flanker radar' it is not unlike saying'Pentium PC' - the range of power aperture product performance seenacross these radars is greater than in any other tier one combataircraft. A detailed analysis of each of the main radar types used isprovided in the latter part of this paper. FlankerRadar Variants RadarType AircraftType RegionalOperators N001 Myech KnAAPOSu-27S,Su-27SK, Irkut/KnAAPO Su-27UBK/PU,Irkut Su-30K Russia,China, Indonesia, Vietnam N001VE KnAAPOSu-30MKK, Su-30MKV China,India, Vietnam N001VEP KnAAPOSu-30MK2 China N001VE-Pero KnAAPOSu-30MK3 TBD N011 KnAAPOSu-27K/Su-35 Russia N011MBARS IrkutSu-30MKI/MKM India,Malaysia N011MBARS IrkutSu-30MKI/MKM India,Malaysia Irbis-E KnAAPOSu-35BM, Su35-1 Russia,Export There are four primaryand distinct types of radar carried by Flankers,each with distinct subtypes or block upgrade levels. In regional terms,variants of the baseline N001 series are most commonly seen, with thevastly more capable N011M BARS phased array now also well establishedin the region. FlankerRadar Power Aperture Product RadarType AntennaDesign AvPower [kW] PAAVE [dBWm2] PkPower [kW] PAPEAK [dBWm2] Range[NMI] 1 m2 RCS LNANF [dB] N001 Myech TwistedCassegrain MSA 1.0 28.0 4.0 34.1 43.0- 53.0 unspec ~9.0 N001VE TwistedCassegrain MSA 1.0 28.0 4.0 34.1 72.0- 81.0 unspec ~9.0 N001VEP TwistedCassegrain MSA 1.0 28.0 4.0 34.1 72.0- 81.0 unspec ~9.0 N001VE-Pero SpaceFeed PESA 1.0 29.4 4.0 35.4 ~102.0unspec ~9.0 N011 PlanarArray MSA 2.0 31.1 8.0 37.1 75.0unspec ~9.0 N011MBARS HybridESA 1.2 28.9 4.8 34.9 ~104.0/ 75.0 spec 3.0 N011MBARS HybridESA 1.6 30.1 6.5 36.2 ~117.5 3.0 Irbis-E HybridESA 5.0 35.1 18.6 40.8 ~153.0 3.5 ZhukAS/ASE AESA est16.3 TBD TBD TBD TBD ~3.0 ComparativePerformance APG-71(F-14D) PlanarArray MSA 0.9 m 7.0 36.5 10.2 38.1 115- 400.0 unspec ~9.0 Notes: Russianspecificationsalmostuniformly assume a 4:1 (25%) duty cycle between average and peak powerratings. Alllater antennaaperturesare0.9meter diameter thus contributing significantly to power apertureproduct. Public specs claim 34 to 36 dB gain for most variants,parametric analysis suggests up to 39 dB gain. Area is ~0.64 squaremetres (0.87 for Pero). In the interests of analytical robustness, thispaper uses the 'raw' form of the peak/average power aperture product(raw power x area [dBWm2]) as this is less vulnerable toincremental design changes in antennas, feeds, and transmitters, thanthe alternate form (power x gain [dBW]. As such the 'raw' form is abetter indicator of the ultimate performance potential of a givenantenna / transmitter configuration. An example of where the latterform [dBW] gets into difficulties is in different specifications ofantenna gain, including or excluding feed losses, or situations wherean incremental design change to a feed or antenna removes several dB ofloss previously listed. Differentclients are often supplied with different TWT ratings on agiven radar type. Detectionrange performance figures for many older radar variants arenot specified in terms of a target Radar Cross Section [RCS] and arethusdifficult to objectively assess. Usually the Russians refer to a'fighter sized target' for such figures, with an assumed RCS of 3 to 5 m2. The muchbetter Noise Figure (NF) performance in the BARS and Irbis is  aresultof the hybrid array design employed. What is clear fromavailable data published by Russian manufacturers over the last fifteenyears is that the Flanker has seen a roughly fivefold growth in thepower aperture product performance of available radar designs. Inpractical terms, this amounts to a gross increase in detection rangeperformance of around 50 percent since the N001 was first deployed, anda gross increase in jammer burnthrough performance of around 120percent. Improvements in processing and antenna/receiver design makethe actual performance improvements better. A good comparison is theAPG-71 radar in the F-14D Super Tomcat, which falls in between theN011M and Irbis E in raw power-aperture product performance (above). Assessing late model Flanker radar range and burnthrough performance byusing the baseline N001 and its variants as a benchmark is therefore acomplete folly. It is akin to using 1980s US fighter radars as abenchmark for assessing the capability of current US production fighterradars. From a strategy and forcestructure planning perspective, the N001 and its variants are now ofmarginal relevance, as the more capable and newer radars willprogressively displace them in the marketplace and thus in deployedforce structures. Very few serious Flanker users will retain N001variants over the coming decade as they are no longer competitiveagainst newer NIIP offerings, or the AESA technology now being deployedby the US and its allies [1].

---

TacticalImplications of High Power Aperture Product Fighter Radars

---

The conventional wisdom in BVR combat is that the player with thelonger ranging radar wins the game as the radar provides theopportunity to detect the opponent earlier, initiate tracking andidentification, and launch a missile shot first. This is howeverpredicated on several assumptions:

1. The player with the longer ranging radar has a longerranging missile to facilitate a first shot.
2. The player with the shorter ranging radar does nothave thecapability to effectively jam the radar.
3. The victim does not turn tail early, spoiling themissileengagement geometry and getting out of the missile's kinematic NoEscape Zone.

The Russian drive to improve supersonic persistence in the Flanker viasupercruise class engines is clearly in a large part driven by thisreality. In the bluntest of terms, throwing a spear from the top of ahill is always easier than throwing one uphill. An R-27EA with a rangecited at ~70 nautical miles becomes a ~100 nautical mile class missileiflaunched supersonic from a superior altitude.

Electronic warfare between opponents remains a key consideration inlong range missile combat. While high power aperture radars providegood burnthrough performance, at extreme ranges well in excess of 50nautical miles burnthrough is unlikely to be a practical proposition.This is because the power ratings of conventional defensive jammingsystems will be sized to defeat surface based engagement radars withpower aperture performance well in excess of any fighter radar. Atechnique for suppressing ajamming source that is  available to users of AESAs and hybridESAs is toput sharp nulls into the antenna mainlobe dynamically.

The use of the ESA or AESA to jam an opponent's radar is a propositiononly where the opponent lacks the frequency agility in the their radarto evade jamming, and lacks an X-band anti-radiation missile whichwouldbenefit from the stable emissions produced by a jamming mode.

What does become a proposition for both sides is jamming of the missilemidcourse datalink uplink channel to deny midcourse flight positionupdates after a missile launch. Historically the jamming of missileuplinks has been considered difficult and demanding of high powerlevels. This is because missile datalink antennas point in thedirection of the launching aircraft, which means that what littlejamming power can couple into the antenna must be carried by surfacetravelling waves along the missile airframe. With a high power apertureESA or AESA such uplink jamming becomes feasible. However, both sidesalso have the option of coating their missiles with X-band lossymaterials, which will diminish the coupling effect.

The reality for better or worse is that possessing radar detection andtracking range performance well in excess of missile kinematic rangeperformance is unlikely to provide any benefit beyond very earlywarning of an inbound threat, giving the pilot the option of reversingand getting away, provided the opponent's radar and radio frequencysurveillance systems are not good enough to detect the longer rangingradar. Increased fighter power aperture performance may increase itstarget detection footprint, but it also increases the opponent'spassive detection footprint for the radar - the inverse square law ofpassive detection produces stronger effect than the inverse fourthpower law of radar detection.

---

Westernvs RussianHighPower Aperture Product Radars

---

The latter phase of the Cold War saw an ongoing contest between US andSoviet designers to deliver the longest ranging multimode radars forBVR combat in intensive jamming environments. The Soviets led duringthe 1960s with the massive RP-25 Smerch on the MiG-25 Foxbat. The USthen gained a lead during the early 1970s with the F-14A's AN/AWG-9radar, originally developed for the navalised F-111B, and the F-15A'sAPG-63 radar. The Russian's snatched the leading position back with theearly 1980s N007 Zaslon on the MiG-31 Foxhound, a massive phased arraydesign twice the size of the US AWG-9 radar. Rated at 2.5 kiloWattsaverage power, with a 25% duty cycle peak power rating of 10 kiloWatts,this immense radar is claimed to be capable of detecting a 0.3 m²RCS cruise missile at 35 nautical miles range.


NIIP's enormous N007Zaslon onthe MiG-31 was the highest power-aperture fighter radar to emergeduring the last decade of the Cold War. If was specifically built tohunt cruise missiles, and is claimed to be able to detect a 0.3 m²RCS cruise missile at 35 nautical miles range.

The advent of the Flanker saw the introduction of the N001 radar,intended to match the US APG-63. While the radar did not meetexpectations, it is the baseline of the Su-30K radar flown by theIndians in the Cope India 2004 exercise in which late model APG-63equipped F-15Cs were defeated in simulated BVR combat.

By the time of Cope India, Russian industry was delivering the firstproduction examples of the NIIP N011M BARS hybrid ESA on early IndianAF Su-30MKI Flankers. Until recently the BARS was the highestperforming radar on any fighter other than the F-22A - while theAWG-9/APG-71 has marginally better power aperture performance, thehybrid array design of the BARS gives it around 6 dB better sensitivity.


The next major advance in the state of the art was the F-22A's APG-77radar, designed for low observability, with the highest (and to datestill undisclosed) power aperture rating of any fighter radar. Itremains the benchmark in this technology, a large 1500 element AESAdesign.

With the APG-77 setting the technological trend, the US industrydeveloped over the last decade a number of AESA upgrades and newdesigns. The APG-79 is the F/A-18E/F Block II radar, originallyintended for all Hornet subtypes but only integrated on the SuperHornet due to its cooling demands. The APG-80 is the F-16/B60 AESA. TheAPG-63(V)2 is a first generation AESA upgrade to the F-15C, theAPG-63(V)3 being a second generation design based on module technologycommon to the APG-79. The APG-81 is the F-35 Joint Strike Fighter AESA,which uses later generation modules than the earliest APG-77 variant.Thecurrent production F-22A/B20 APG-77(V)2 uses common module technologyto the APG-81, but delivers considerably more power due to the largermodule count and greater cooling capacity of the airframe.

The US had a major technologybreakthrough in AESA design around a halfdecade ago when Gallium Nitride (GaN) HEMT (High Electron MobilityTransistor)X-band transistors wereperfected, allowing considerably more output power than earlier GalliumArsenide transistors. This has created an effect not unlike euphoria insome parts of the US defence industry, and a worldwide drive by globalsemiconductor houses to occupy the market. Historically AESAperformance was limited by the power output per module at X-band,typically of the order of 2 to 5 Watts per module. The GaN transistortechnology appears at this stage to be capable of delivering ten timesthe power per module, which changes the problem AESA designers facefrom barely getting viable power output, to not having enough coolingand electrical power capacity to cope with the transistor technologyavailable. A good example is the Toshiba TGI8596-50 GaN HEMT announcedJuly, 2007, capable of delivering 50 Watts in the X-band and targetedatradar and microwave communications equipment.

The long term implications of the Gallium Nitride breakthrough inX-band microwave transistor technology are most interesting. If AESAdesigners are not significantly limited by basic technology in themicrowave power they can extract from each AESA module, then radarpower aperture performance will grow until it hits the limits of thepower generation and especially cooling capacity of an airframe.

Consider a radar design with 1500 modules, and the availability ofmodules capable of, if powered and cooled adequately, transmitting 40Watts of continuous wave X-band microwave power, with an efficiency of50 percent (PAE=50%). The sustained peak power such an AESA couldproduce isof the order of 60 kiloWatts. If we assume the Gallium Nitridetransistors are capable of sustaining 160 Watts each this power ratingcan be quadrupled to P_PEAK=240kiloWatts. For comparison X-band GaN/SiC transistors rated at 80 Wattshave already been reported in the research literature. With an aperturearea of about 0.65 m² thisyields a power aperture product of the order of 51.9 dBW or a relativerange increase compared to contemporary top end 20 kiloWatt classfighter radars of around 150 percent. The utility of this rangeincrease may be irrelevant considering conventional targets, but whereit matters is in providing the ability to detect stealthy targets atvery good ranges. The following chart depicts the impact of a notionalvery high power aperture radar on detection ranges for stealthy targets.


What is clear is that X-band fighter radars with peak power ratingswell above 20 kiloWatts have the potential to render all but top endstealth technology ineffective. While engineering such radars wouldpresent serious challenges, some arguably extremely difficult toresolve with a sub one metre aperture diameter, and possibly forcingvery low operating duty cycles, it is abundantly clearthat the trend willbe to strive for the highest power aperture product achievable, as theincentives are very powerful.

In this game the primaryconstraints then become the cooling of the array and dumping of wasteheat out of the aircraft. Larger aircraft do much better with theseconstraints, compared to smaller aircraft. In the long term contest forhigher  power aperture product, fighters like the F-35 JSF,F/A-18E/F, F-16 cannot compete with aircraft in the size and volumeclass of the F-15, Flanker or F-22A. The defining characteristics forbest survivability will be the size to effectively power and cool thehighest power aperture product radar which can be fitted, and the bestX-band all aspect stealth performance.

The potential of X-band fighter radars with power ratingsin excess of 20 kiloWatts to be used as Directed Energy Weapons (DEW)is an issue in its own right [2].

The Russian response to the surge in US AESA production was to launchthe development of the 20 kiloWatt peak power class Irbis E radar, anevolution of the N011M BARS. This radar is to be carried by the newSu-35S (formerly Su-35BM and Su-35-1) Flanker variant.



Of the current generation of US AESA radars, the only one which is welltechnically documented in the open literature is the APG-79, which willtherefore be used as a baseline for comparison against the Flankerradars.

The APG-79 was initially sold as a block upgrade to the legacy APG-73,itself an incremental upgrade to the APG-65. The APG-79 however endedup being much more than a simple block upgrade, adding not only apowerful AESA, but including additional processing capability and tightintegration with the ALR-67 radar warning and emitter locating system,and requiring forward fuselage changes to the aircraft.One of the key design considerations was to improve the capability todetect and engage anti-shipping cruise missiles, a major problem forthe US Navy Carrier Battle Groups. Given the relatively modestfootprint to be defended, the poor supersonic performance and payloadrange of the Super Hornet was less important than the ability to liftan X-band radar above the horizon of the shipboard defences.

There is enough unclassified data available at this time to perform areasonable estimation of performance bounds on this radar, with thecaveatthat evolving transistor technology over the life cycle of the designwill see shifts in performance. The radar is known to have ~1100modules, which assuming like per module power rating, cooling andX-band wavelength would result in around 70 percent of the power ratingof the APG-77. This puts the radar broadly between 10 kW and 20 kW peakpower ratings. Public data comparing the APG-71, APG-73 and APG-79yields an indication that the radar has similar power aperture productperformance to the 10 kW rated APG-71, which for half the antenna areayields a peak power rating of the order of 20 kW. This data supportsthe proposition that the radar is a 20 kW peak power class design.

In general, the peak power rating of an AESA is determined by the permodule power rating multiplied by the number of elements, with somereduction resulting from the taper function which is used to weightpower output per module, so that sidelobes and mainlobe shape can beoptimised. A 20 kW peak power AESA with a 15% allowance for taperfunction yields for instance a per module rating, for 1100 modules, ofaround 21 Watts. The average power output of the radar is then limitedby the duty cycle of operation, and power consumption overheadsincurredby drivers, and phase and control elements in the modules.

The latest engineering literature on AESAs puts the state of the artfor radiated X-band power intensity at about 4 Watts/cm²which for the X-band is around 16 Watts/module. This would put thetotal peak power at about 17.6 kW.


This chart shows asampling ofperformance figures from recent research papers for GaAs and GaNtechnology Monolithic Microwave Integrated Circuit (MMIC) High PowerAmplifiers (HPA)operating at X-band frequencies. Such devices are the primary powerstage in an AESA Transmit-Receive module. GaN HEMT technology hasdriven up recent power ratings to better than 20 Watts, butefficiencies remain problematic, mostly below 40%.

Prior to the advent of the GaN HEMT, the conventional wisdom aboutAESAs was “AESAs are great for high average power but not so great forpeak power”, reflecting limited per transistor power ratings. Clearlythe state of the art currently permits a 20 Watt module, using eitherganged HEMTs of lower power, or single GaN HEMTs delivering all of thepower, the principal question then being whether the PAE (efficiency)is at the high or low end of the scale.

These estimates can be further constrained by applying someunderstanding ofbasic AESA design principles, and constraints such as the publiclydisclosedcoolingdemand of the AESA antenna section, which dominatesthe PAO liquid cooling loop load.  Very little modelling isrequired to relate the waste heat dissipation to cardinal performanceparameters of the HEMT power transistors, and factors such as typicaltransmit duty cycles [3].

However, accuracy of estimation depends on the assumed mode in whichthe HEMTs in the TR modules are operated. Are they running in staticA-class operation, or are they running in pulsed or gated A-class mode,where the transistors when idle are biased down to a low powerconsumption mode? Another question is how many Watts of power can beeffectively extracted from each module by the cooling system? Anotherconsideration is the total duty cycle of the radar, or the percentageof time it is transmitting - while individual modes may occupy up to25% duty cycle each, interleaving multiple modes will drive up the dutycycle considerably, in turn driving up average power dissipation. Afurther consideration is what static thermal load may be incurred byother components of the radar, such as processors, if these are liquidcooled half a kiloWatt or more may be required to cool a full cardcageof VMEbus COTS processors. Modelling this, assuming 0.5 kW for non-AESAheat load, a 20% power overhead for the AESA internal support circuitsand backplane driver amplifiers, and a per module efficiency between25% and 45% shows that peak power ratings of the order of 20 kW arefeasible at duty cycles of the order of 50%, with acceptable coolingmargin.

The available data and modelling thus easily supports the propositionthat theAPG-79 is a 20 kW peak power class radar, with the caveat that betterperformance is theoretically achievable, as is lower performance giventhe technology available. With around one half the aperture area of theBARS and Irbis E radars, the resulting order of magnitude poweraperture product of 38 dBWm² puts the radar almost exactlyin between the twoRussiandesigns. If we are generous with assumptions and consider growthpotential, then it might be rated a little closer to the Irbis E.

It is also worth asking the question of what peak power rating theAPG-79 would require to match the 40.8 dBWm² power apertureproduct ofthe Irbis E, given that its aperture is around half the area of theFlanker radar. The result is a considerable 35 kW - reflecting thereality that half the antenna size requires twice the peak power tomatch a power aperture figure. Is such performance feasible? The pertransistor module rating is then 32 Watts, which is feasible butquite challenging. The peak radiant power density at the face of theantenna is 8 W/cm² which is around twice the cited currentstate of the art. The HEMTs would have to be state of the art, butbasically such peak power performance ispushing against the capacity of the extant cooling system, andachievable transistor performance in efficiency and power output.

Claims that the APG-79 canoutrange the Irbis E are very difficult to support given basic radarphysics. A claim of a tactically significant range advantage over theextant BARS is also hard to support. The corollary of this is that alate model F-35 Joint Strike Fighter APG-81 with similar module count,module power andaperture size to the APG-79 will not provide significantly differentperformance, relative to the later Russian radars.

This chart comparesgraphicallythe peak power aperture product estimates for the Flanker radars, andthe APG-79. The latter is depicted with some provision for growth.Hybrid array technology used in the BARS and Irbis E provide similartotal noise figure for the antenna-receiver design to that of AESAdesigns like the APG-79. Growth JSF APG-81 radars will be similar tothe APG-79. Block upgrades to the BARS to convert it into an Irbis Econfiguration will not present difficulties as the latter is an upratedderivative of the former.

The N011 radar was a higherperformance replacement for the lacklustreN001 series. It is distinctive by its use of a backplane fed planararray antenna, a design very similar to the US Hughes and WestinghouseAPG-6X series radars which emerged during the 1970s and are only nowbeing replaced by newer AESAs. The N011 uses largely digitalprocessing, unlike the hybrid N001 series. An L-band IFF interrogatorantenna array is embedded in the X-band planar array.

The N011 was only ever built in modest numbers, to equip the RussianAir Force Su-27K, later renamed Su-35. It most closely compares to theAPG-63 and APG-70. By the mid 1990s Russian AF interest shifted to theN011M BARS phased array, trialled in the Su-37 demonstrators.

The Zhuk series was developedfor the stillborn mid 1980s MiG-29M/MiG-33 Fulcrum upgrade andproduction effort. Designated the N010 Zhuk, this was a relativelymodern pulse Doppler design modelled on the US APG-65 and APG-68radars, using a slotted planar array antenna with a 0.68 metre diameteraperture, with an average power rating of 1 kW and peak rating of 5 kW.With the end of the Cold War and Phazotron's emergence as anindependent entity in an open market, the effort invested into the Zhukwas exploited to develop a family of radars designed for the MiG-29,Su-27/30 and older Soviet era fighters as upgrades.

The Zhuk-27 was a variant of the baseline N010 but fitted with a muchlarger 0.98 metre diameter slotted planar array antenna, and possiblyan uprated TWT, intended for the Su-27SK Flanker B. Its contemporarywas the Zhuk-8P developed for the PLA-AF J-8-II Finback, with a smallerantenna and thus lower range performance.

The baseline mechanically steered Zhuk further evolved, with the N010MZhuk-M and Zhuk-ME variants for the MiG-29 Fulcrum, and Zhuk-MS andZhuk-MSE intended for the Su-27/30 Flankers. These incorporated anarray of L-band IFF dipoles, a slotted planar array, and much improvedprocessor hardware, to support strike modes including SyntheticAperture Radar imaging.

The new slotted planar array radar installed in the Shenyang J-10BSino-Flanker bears a close resemblance to the Zhuk-MSE and may be alicenced or reverse engineered variant of this design.

The Zhuk MSF/MSFE (above) is a passive ESAdesign intended to compete against the NIIP N011M BARS. It uses aPhazotron unique radial distribution arrangement in the backplanewaveguide feed, and proprietary radiating element placement. The ZhukMSFE has a .98 meter diameter aperture with 1662 radiating elements,and was developed for the Su-30MK3 Flanker G avionic suite intended forthe PLA-AF (MAKS 2005).

The most advanced of the Phazotron Flanker radars is the Zhuk-MSFE PESAvariant, currently being flight tested on the Su-27KUB/Su-33UBside-by-side cockpit navalised Flanker variant, likely to be acquiredby the PLA-N as part of their intended carrier airwing for the VaryagCVA.

This radar is usually credited with a 2 KW average power rating and 8kW peak power rating, putting it in the performance class of the NIIPN011 MSA radar on the Su-27K/Su-35 Flanker E. The PESA design has 1662radiating elements.


The Zhuk-MSFE is beingflown inan Su-33UB demonstrator, the depicted example (below) with thrustvectoring Al-31FU engines (MAKS 2007).

The BARS is the mostadvancedradar developed by Russian industry during the 1990s. It is unusual inbeing designed with a hybrid array arrangement, the receive path usingvery similar technology to US and EU AESAs, with similar sensitivityand sidelobe performance, but using a Travelling WaveTube and backplane waveguide feed for the transmit direction, atechnology closest to the B-1B and early Rafale EA radars. As such theBARS is a transitional design sitting in between Passive ESAs (PESA)andcontemporary AESAs. There is no doubt this design strategy reflectedthe unavailability to Russian designers of the Gallium Arsenide powertransistors used in Western AESAs.

The baseline N011M radar uses avertically polarised 0.9 metre diameter aperture hybrid phased array,with individual per element receive path low noise amplifiersdelivering a noise figure cited at 3 dB, similar to an AESA. Theantenna is constructed using phase shifter and receiver 'stick'modules, a similar technology to early US AESAs.

Threereceiver channels are used, one presumably for sidelobe blanking andECCM. The EGSP-6A transmitter uses a single Chelnok Travelling WaveTube, available in variants with peak power ratings between 4 and 7kiloWatts, and CW illumination at 1 kW. Cited detection range for aclosing target (High PRF) is up to 76 NMI, for a receding target up to50 NMI. The phased array can electronically steer the mainlobe through+/-70 degrees in azimuth and +/-40 degrees in elevation. The wholearray can be further steered mechanically. Polarisation can be switchedby 90 degrees for surface search modes.

The BARS remains in production for the Indian and Malaysian Irkut builtSu-30MKI/MKM variants. The radar is available with a range of TWT powerratings, this being the source of considerable confusion to observerswho have not tracked this program since its inception. The result is awide range of performance figures depending on the resulting PowerAperture Product. That the antenna has good power handling capabilityis evident in its adaptation for the Irbis E design.

Given the similarity between the Irbis E and BARS,existing BARS operators will over time effect block upgrades to converttheir BARS inventories into the Irbis E configuration.

In 2009 there were twoprincipal candidate AESAs for installation in new build Flankers, orretrofit into existing service Flankers. These radars are NIIRPhazotron's intended Zhuk-AS/ASE, scaled up from the MiG-35 Zhuk AEAESA, and a derivative of Tikhomirov NIIP's new PAK-FA AESA, displayedpublicly at MAKS 2009.

Both radar designs are based on the quad channel TR module technologyfirst disclosed during the public release of the Zhuk AE. These X-bandmodules are now being mass produced on an automated line by NPP Istok,who are also planning S-band module production. Mostly Russian producedGaAs components are employed. Cited capacity is sufficient for 50 AESAradars annually.

Other than a stated intent by NIIR Phazotron to scale up the Zhuk AE,there are no technical details of this design available at this time.In a sense it is an analogue of the Raytheon scaling of the APG-79 AESAfor the APG-63(V)3/4 upgrade (refer below).

Estimated detectionrange chartfor variants of a Flanker sized AESA equipped with a range of TransmitReceive Module power ratings per channel. The detection rangeperformance of the 10 and 12 Watt module equipped AESA is similar tothe Tikhomirov NIIP Irbis-E hybrid ESA in the Su-35S, and much superiorto the N011M BARS. The performance of AESA if equipped with modulesrated above 15 Watts is superior to the Irbis E. Receiver noise figureand effective aperture area are assumed to be similar. N011Mperformance is based on parametric data and is better than NIIP citedfigures  (Author, 2008).

Modelling performed by APA in 2008, making some reasonable assumptions,such as an element count of ~1600 for the antenna provides a goodbaseline for a Zhuk AS/ASE as well as the PAK-FA AESA. This is detailedunder theZhuk AS/ASE analysis of2008.



Enhanced stills from aRussiantelevision broadcast reporting the Tikhomirov NIIP PAK-FA AESA design.Static display images of the antenna have a dielectric impedancematching screen installed, which obscures the actual TR moduleapertures (Vesti - Moskva via Youtube).


The Tikhomirov NIIP AESA design for the PAK-FA is better understood.The antenna aperture is very similar in size, if not identical, to theaperture of the Irbis E. The design is intended for fixed low signaturetilted installation, rather than gimballed installation, and auxiliarycheekarrays are planned for. The design is claimed to have been integratedwith an existing BARS/Irbis radar for testing and design validationpurposes.

Public statements made in Russia claim 1,500 TR module elements.Counting exposed radiating elements on video stills of the antennaindicates an estimated 1,524 TR channels, with a tolerance of severalpercent. This is within 5% of the 2008 APA model for a Flanker AESA.

NIIP have publicly cited detection range performance of 350 to 400 km(190 to 215 NMI), which assuming a Russian industry standard 2.5m²target, is consistent with the 2008 APA model for a radar using ~10Wrated TR modules, which in turn is the power rating for the modulesused in the Zhuk AE prototypes. This puts the nett peak power at ~15kiloWatts, slightly below the Irbis E, but even a very modest 25%increase in TRmodule output rating would overcome this.

There are distinct differences between the AESA displayed by NIIP forVesti, which has less depth and uses circular radiators, and theexamples displayed at MAKS 2009 and depicted on brochures, which areconstructed using TR module sticks and are several inches deeper.

To drive down the cost of this AESA, the best strategyavailable to the Russians is the export of AESA upgrades to the globalcommunity of Flanker users over the coming decade, emulating the USapproach with this technology. Tikhomirov NIIP brochures state that theexisting AESA would be the basis of AESA upgrade designs for theSu-27/30/35 Flankers.

Recent reports from India suggest that 100 Su-30MKI Flanker H may beretrofitted with AESA upgrades to their N011M BARS radars post 2015.

It is now inevitable that AESAs will appear on Flankers, the onlyuncertainties at this stage will be in the number of aircraftretrofitted, the clientele, and exact timelines and performancespecifications of these radars.

[1]There are numerousreports of PLA dissatisfaction with the N001 series radars, supportedby reports of the Pero demonstrator sale to China. This presents thelikely outcome of the PLA-AF acquiring the Irbis E equipped Su-35, butalso performing block upgrades to the extant Su-27SK/SMK and Su-30MKKfleets as immediate force structure expansion costs taper off after2010.
[2] Refer C Kopp,Considerations on the use ofairborne X-band radar as a microwave directed-energy weapon,Journal of Battlefield Technology, vol 10, issue 3, Argos Press PtyLtd, Australia, pp. 19-25.
[3] A major factor is the achievable performance of theHEMT transistors installed in the AESA Transmit Receive modules, ietheir X-band power rating in Watts, and their Power Added Efficiency(PAE) in percent, the latter a measure of excess waste heat dissipationin the modules. The high bandwidth and linearity demands imposed onmilitary AESAs generally force the use of A-class amplifier designs,which are profligate consumers of electrical power and thus heatdissipators, regardless of clever power management techniques. Thecurrent state of the art in X-band HEMT transistors and MMIC(Monolithic Microwave Integrated Circuits) sees PAE values ranging from25% to the the nominal 45% up to almost 70% in pulsed modes, and powerratings fromsingle Watts up to a staggering 80 Watts per transistor reported in anacademic journal (Toshiba). No matter how good the transistor might be,the hard limits on average and thus total power will be set by thecapacityof the liquid cooling system.
[4] Chinese interest in the Pero may well be driven byan imperative to increase capability at minimum cost. While it is knownthat NIIP have tried to market the BARS for block upgrades, reportsindicate a reluctance on the part of the PLA to embrace a system whichis identical to what the Indians are using. A Pero block upgrade with astrongly uprated transmitter is the cheapest path for the PLA to matchor exceed the PA of the BARS.

   

Related Air Defence Materials:

SAM System Index - Missile Systems vsEngagement and Acquisition Radars [Click for more...]

SAMSystem Mobility - Air Defence SystemVehicles [Click for more ...]

SAM SystemIntegration - Air Defence Command Posts [Click for more ...]

SAM System Passive Targeting -Emitter Locating Systems [Click for more ...]

SAM System CounterVLO Capabilities [Clickfor more ...]

SAM SystemProliferation - PLA AirDefence Missile Systems [Click for more ...]

SAM System Multimedia -RosoboronexportFootage [Click for more ...]