Radars for Ballistic Missile Defense Research · • INGWERSEN AND LEMNIOS Radars for Ballistic...

• INGWERSEN AND LEMNIOSRadars for Ballistic Missile Defense Research

VOLUME 12, NUMBER 2, 2000 LINCOLN LABORATORY JOURNAL 245

Radars for Ballistic MissileDefense ResearchPhilip A. Ingwersen and William Z. Lemnios

■ Lincoln Laboratory’s involvement in ballistic missile defense began over fortyyears ago at the time of the first launches of ICBMs and satellites by the SovietUnion and the formation of the Advanced Research Projects Agency (ARPA).The Reentry Physics Program, started in 1958 and sponsored by ARPA, set outto understand the behavior of hypervelocity objects reentering the atmosphere,with the expectation that this research would lead to a means of discriminatingbetween warheads and decoys. The program, which combined theoreticalanalysis, laboratory experiments, and field measurements, provided a foundationthat soon led to other similar programs. The U.S. Air Force, interested in theperformance of its own ICBMs against enemy defense systems, also initiated aprogram of radar development and measurements similar to that of ARPA. As aconsequence, the Laboratory became heavily involved in ARPA’s Project PRESS(Pacific Range Electromagnetic Signature Studies) and ARPAT (ARPATerminal) programs and the Air Force Penetration Aids program. By 1963 thesethree large programs, combined with related efforts in the development of radartechnology, occupied approximately half of Lincoln Laboratory’s staff. Fifteenlarge sensitive radars designed for signature measurements were built as a result,and Lincoln Laboratory had some role in the development of each. This articletraces the history of the measurement radars and the technology programs thatsupported them. It concentrates on the four major radars at the KwajaleinMissile Range. These radars continue to play a major role in the development ofballistic-missile-defense systems and discrimination techniques.

I in what was thencalled AICBM began at Lincoln Laboratory.The initial goal was to study the detection and

tracking of enemy intercontinental ballistic missiles(ICBM) in order to gain early warning of an enemyattack. The Millstone radar (see the article entitled“Radars for the Detection and Tracking of BallisticMissiles, Satellites, and Planets,” by Melvin L. Stoneand Gerald P. Banner, in this issue) was also started in1956 as a prototype of an early-warning radar. Twoevents in 1957 resulted in the acceleration of effortsby the United States to develop systems capable ofcountering attacks by enemy ICBMs. First, on 26 Au-gust 1957, the Soviet Union announced that it had

successfully tested an ICBM. Second, on 4 October1957, the world was shocked to hear the beeping ofSputnik I, the first artificial satellite. The Departmentof Defense reacted swiftly to these two events andformed the Advanced Research Projects Agency(ARPA) in early 1958. ARPA was given broad juris-diction over research and development of spaceprojects and antimissile systems. ARPA’s principalprogram, called DEFENDER, focused on solvingwhat was perceived to be the most difficult antimissileproblem, namely, discrimination between warheadsand the debris and countermeasures expected to ac-company the warhead.

In the vacuum of space, objects such as chaff, bal-


246 LINCOLN LABORATORY JOURNAL VOLUME 12, NUMBER 2, 2000

loon-like replica decoys, and fragments of the launchvehicle move at the same velocity as the warhead.Chaff can hide the warhead, while balloon-like rep-lica decoys and fragments of the launch vehicle canhave radar signatures sufficiently similar to that of thewarhead to make discrimination quite difficult.Tracking all these objects and assigning interceptorsto them when discrimination is not possible can eas-ily exhaust the radar and interceptor resources of adefense system. As atmospheric drag is encounteredin reentry, the lighter objects slow down and fall backto reveal the threatening heavier vehicles, includingthe warheads. The lower the altitude at which theslowdown occurs, the less time there is for defense-system reaction and interceptor launch and fly-out.Thus one of the more stressing countermeasures forthe defense system to overcome is the reentry decoy, asmall, heavy object that can match the radar cross sec-tion and deceleration characteristics of the warheadall the way down to low altitudes. Learning how toperform discrimination between warheads and reen-try decoys at as high an altitude as possible was a ma-jor goal of ARPA and the DEFENDER program.

Hypervelocity objects encounter increasing drag as

they enter the atmosphere. The kinetic energy theylose heats and ionizes the surrounding air, leaving anionized trail. From studies of meteors, scientists knewthat this ionized wake could be detected at radarwavelengths as well as in the visible spectrum. Theyreasoned that the energy of the wake should be relatedto the mass of the object that produced it, and thusradar measurements of the ionized wake should yielda means of discriminating between the heavy war-heads and the lighter reentry decoys. Building radarsthat had the sensitivity, coherence, and resolution inboth range and range rate to make the required mea-surements was a challenging task that required manyadvances in the state of the art. The DEFENDERprogram set out to solve the problems, and LincolnLaboratory was given the opportunity to work onmany of them.

Reentry Physics Program

In July 1958 the ARPA-sponsored Reentry Physicsprogram started at the Laboratory under the technicaldirection of Daniel E. Dustin, then a group leader,and his assistants, Glen F. Pippert and Leo J. Sullivan.The lofty goal of the program was “to determine theeffects of the ionization produced by a reenteringbody on the electromagnetic scattering characteristicsof the body,” and “to develop adequate theoreticalmodels to explain the experimentally observed phe-

FIGURE 2. The experimental chamber of the RSR was largeenough to accommodate multiple optical and microwavemeasurement instruments. The principal instruments wereschlieren cameras and a cavity resonant at UHF, which pro-vided quantitative measurements of electron density andwake diameter.

FIGURE 1. The two expansion chambers and the experimen-tal chamber (in the rear) of the Reentry Simulating Range(RSR) constructed by Lincoln Laboratory in 1960. This facil-ity utilized high-velocity pellets to simulate and measure thephysics of atmospheric reentry. Pellets entering from theright in the picture pass through the two cylindrical expan-sion tanks in the foreground to dissipate the combustionproducts of the gun before entering the experimental cham-ber in the background. This large system could be evacu-ated to a pressure of fifty microns of mercury in only fiveminutes.



nomena” [1]. Laboratory measurements were to bemade on small pellets at a number of hypervelocitytest facilities, including one to be built at LincolnLaboratory. Field measurements of somewhat largerreentry objects would be provided by radars atArbuckle Neck, Virginia, observing vehicles launchedfrom Wallops Island, Virginia.

Reentry Simulating Range

In 1960 a Reentry Simulating Range (RSR) was con-structed near the Laboratory and equipped with aconventional powder gun that could fire 0.5-inch-di-ameter projectiles at velocities up to 9200 ft/sec, anda light-gas gun that could fire 0.186-inch-diameterspherical pellets at velocities of over 20,000 ft/sec.The light-gas gun fired a three-inch-diameter projec-tile to compress hydrogen in a barrel blocked with afrangible window. When the window broke, the com-pressed gas was released into a smaller barrel, where itaccelerated the pellet to ICBM velocities. Figure 1shows the expansion tanks and experimental chamberof the RSR; Figures 2 and 3 show the experimentalchamber in more detail. Figure 4 shows the light-gasgun that fired pellets at ICBM velocities, and Figure 5shows some of the pellets that were used by both ofthe guns.

The RSR experimental chamber accommodatedmultiple optical and microwave measurement instru-ments; schlieren cameras and a resonant UHF cavity

were the principal instruments [2]. Figure 6 shows anexample of the photographs made with these instru-ments. These photographs provided an excellentqualitative picture of the projectile’s flight, the bowshock wave, the growth of the wake diameter as afunction of distance behind the object, and the loca-tion of the point at which the wake flow changedfrom laminar to turbulent. The resonant UHF cavityprovided a time history of the pellet’s passage, andthus measurements of the atmospheric ionization asfunction of distance behind the body. Amplitude andphase measurements could be interpreted in terms ofwake diameter and electron density, and changes inthe wake’s growth rate were used to determine the po-sition of the transition from laminar to turbulentflow. Because of its ability to perform a large numberof experiments in a relatively short period, the RSRproved invaluable for determining the effects of pres-sure, velocity, and materials on the wake properties. Itwas used until 1970.

Arbuckle Neck Radars

Lincoln Laboratory’s first two reentry-measurementradars were designed and built at the Laboratory in1959 and installed at Arbuckle Neck, Virginia, to ob-serve Trailblazer I and Trailblazer II vehicles launchedby NASA from Wallops Island, Virginia [3].

The Trailblazer I was a six-stage missile that usedthree stages to boost its velocity package to an altitude

FIGURE 4. The light-gas gun used to fire pellets at ICBM ve-locities. This gun routinely achieved pellet velocities of over20,000 ft/sec. On one occasion a hydrogen leak caused a pel-let to be fired at 32,000 ft/sec.

FIGURE 3. The inside of the RSR experimental chamber isshown with an L-band waveguide experiment that was usedto measure the transmission and reflection properties of theionized medium. The experimenter is Melvin A. Herlin, theleader of the group responsible for the RSR.



of approximately two hundred miles and the remain-ing three stages to accelerate the payload back towardWallops Island. The system achieved a reentry veloc-ity of 20,000 ft/sec with a two-pound payload. Be-tween March 1959 and July 1962, fourteen of theTrailblazer I vehicles were flown, the majority withfive-inch spherical reentry bodies of various materialsand coatings. In order to fly larger vehicles, more rep-resentative of warheads, NASA designed the Trail-blazer II, which was a four-stage rocket that couldachieve a velocity of 20,000 ft/sec with a thirty-five-pound payload. Firings commenced in 1963 andcontinued through the end of the Laboratory’s opera-tions at the site in 1965.

The first of the two reentry-measurement radars atArbuckle Neck, the S-band tracker, was built in alittle over one year, in time to successfully track the1 December 1959 launch of the third Trailblazer I. Tomeet the accelerated schedule, Laboratory researchersassembled the system primarily from available parts.The antenna was a sixty-foot dish on a surplus U.S.Navy five-inch gun mount. The transmitter was froman AN/FPS-6 radar and utilized an S-band magne-tron capable of 4.5-MW peak power at a pulse widthof 2 µsec. Tracking hardware was from an SCR-584, aWorld War II fire-control radar for antiaircraft guns.The system used conical-scan angle tracking and asingle receiver channel with a cooled parametric am-

plifier designed and built at Lincoln Laboratory. Theconical scan was used because at that time it was notpossible to build parametric amplifiers with the sta-bility required for a three-channel monopulse system.

The second radar at Arbuckle Neck, also utilizing asurplus U.S. Navy gun mount and a sixty-foot dish,had diplexed UHF and X-band systems that wereslaved to the S-band tracker to form the first inte-grated multiwavelength data-gathering system. With

FIGURE 7. Radar antennas at the Lincoln Laboratory fieldsite at Arbuckle Neck, Virginia, located near the Wallops Is-land, Virginia, launch facility. Site instrumentation consistsof (from right to left) the S-band tracking radar, the multi-plexed UHF and X-band cross-section measurements radar,and the SPANDAR (Space Range Radar) long-range trajec-tory and range-safety radar designed and built by LincolnLaboratory for NASA.

FIGURE 6. A schlieren photograph of a 12.5° half-angle conetravelling at 5700 ft/sec. The photographs provided a qualita-tive measure of reentry events, which aided the interpreta-tion of the more quantitative microwave measurements. Thepointer in the image indicates the approximate position ofthe transition from laminar to turbulent flow.

FIGURE 5. Assorted pellets used at the RSR. The 0.5-inchprojectiles, including all the non-spherical objects, could befired only by the conventional powder gun. The sphericalpellets smaller than 0.5 inches were fired only by the higher-velocity light-gas gun. True ICBM velocities were achievedonly with the 0.186-inch-diameter pellets.

Transition to turbulent flow



little knowledge of what wake cross section to expect,the designers built the UHF system to transmit alter-nating 1-µsec and 6-µsec pulses. The transmitter useda version of the VA-812 klystron developed for theBoston Hill radar (see the article entitled “Long-Range UHF Radar for Ground Control of AirborneInterceptors,” by William W. Ward and F. RobertNaka, in this issue) and used again for the ARPALong Range Tracking and Instrumentation Radar(ALTAIR). Although the transmitter chain was coher-ent, only amplitude detection was used. The systemtransmitted a vertically polarized signal and receivedboth linear polarizations. The X-band system had theworld’s first X-band maser preamplifier, which wasdesigned and built at Lincoln Laboratory. The carefulwording of Table 1, copied from the first SemiannualTechnical Summary Report of the Reentry PhysicsProgram [1], indicates the uncertainty about the per-

formance that would be achieved by the low-noisepreamplifiers. The report also raised concerns aboutthe feasibility of making a sixty-foot paraboloid ofsufficient accuracy for operation at X-band. In theend, only the central portion of the dish was used atX-band in order to have a broad enough beam to keepthe target illuminated in the slaved mode of opera-tion. Figure 7 shows the Lincoln Laboratory S-bandand UHF/X-band radars built for ARPA at ArbuckleNeck, and a third system called SPANDAR (SpaceRange Radar) that was built by Lincoln Laboratoryfor NASA. SPANDAR, which operated at S-band,also had a sixty-foot dish, but used the superiormount design of the Millstone radar. It was designedand used as a range-safety radar for missile launchesand for satellite tracking.

The emphasis on digital recording, which becamea hallmark of the Lincoln Laboratory measurement

Table 1. Approximate System Parameters for the Reentry-MeasurementRadars at Arbuckle Neck, Virginia

UHF S-band X-band

Frequency (MHz) 400 2800 9000

Output power (peak in MW) 8 5 1

Pulse width (µsec) 6 2 2

Pulse-repetition frequency (Hz) 320 320 320

Antenna gain (dB) 35 53 60

Beamwidth (degrees) 2.9 0.3 0.12

Reflector diameter (ft) 60 60 60

Antenna efficiency assumed (percent) 50 50 40

Intermediate-frequency signal-to-noise ratio for a five-inch sphere at a distance of 200 nautical miles,neglecting losses.

Using conventional receivers (dB): 8 13 10

Improvement expected withlow-noise preamplifier (dB): 4* with 10* with 10* with

parametric parametric maseramplifier amplifier

* These approximate signal-to-noise ratios are on a per-pulse basis with no integration. A systemnoise temperature of 300K, corresponding to a conventional noise figure of 2 (or 3 dB), is assumedachievable at each frequency with the parametric amplifier or the maser.



radars, and the practice of recording digital data atfield sites and bringing it back to Lexington for analy-sis, were first established with these sensors. In thiscase, the data consisted of time, range, azimuth, andelevation angle, and one range-gated amplitude sam-ple from each pulse.

Much reentry phenomenology, the knowledge ofwhich we now take for granted, was first observed bythese radars and came as a surprise. The cross-sectiondip that occurs in early reentry, the very large crosssection of the turbulent wake, and the large cross sec-tion of the orthogonally polarized return are ex-amples. The Millstone radar in Westford, Massachu-setts, some seven hundred kilometers from ArbuckleNeck, also observed the Trailblazer launches and pro-vided the coherent data that first showed how rapidlythe mean wake velocity decays. In 1965, with data onfull-scale vehicles available from the Target Resolu-tion and Discrimination Experiment (TRADEX) ra-dar at the Kwajalein Atoll in the Pacific Ocean, Lin-coln Laboratory ceased operations at Arbuckle Neckand transferred the radars to NASA. For more detailson the Arbuckle Neck sensors, the vehicles, and theresults, see the article by Sullivan in the summer 1991issue of the Lincoln Laboratory Journal [3].

ARPA Radar Technology Program

In July 1959 a second ARPA-sponsored effort startedat Lincoln Laboratory. Called the Radar TechniquesStudy, the program had an extremely broad workstatement designed “to allow researchers to developideas with a minimum of direction” [4]. Topics in-cluded, but were not limited to, high-power radio-frequency (RF) sources, high-resolution techniques,improved radar accuracy, low-noise receivers, and ex-traction and utilization of information in radar sig-nals. Detection of signals in non-Gaussian noise andradiometry were two of the initial areas studied. In1960 a larger effort called Microwave Power wasstarted at the Laboratory for ARPA. The tasks in-cluded the design of high-power duplexers and basicwork on vacuum voltage breakdown and electron op-tics, which was needed to design improved RF ampli-fiers and modulator switch tubes. The first computercode to model the saturated gain of a klystron wasone of the results.

U.S. Air Force Interests

The U.S. Air Force, interested in how U.S. ICBMswould perform against an enemy defense system,wanted to minimize the cross sections of their war-heads and to develop decoys and other penetrationaids. To test their systems, they needed the same typeof measurement radars and analysis techniques thatARPA was developing. The Air Force planned to testICBMs in the Atlantic by using sensors on AscensionIsland and on the Mobile Atlantic Range Station(MARS) ships. During 1960 both ARPA and the AirForce were holding discussions with Lincoln Labora-tory about enlarged programs. On 29 August 1960the Director of Defense Research and Engineering di-rected ARPA and the military services to execute asingle integrated program in penetration aids, targetidentification, and reentry physics. Lincoln Labora-tory was requested to provide scientific direction. As aresult, the Laboratory assumed responsibility forARPA’s Project PRESS (Pacific Range Electromag-netic Signature Studies), and a multifaceted Air Forceprogram called Penetration Aids was initiated underthe leadership of V. Alexander Nedzel. The scope ofthe program included the following three elements:(1) participation, with responsible Air Force agenciesand their contractors, in preparing general conceptsand detailed plans for testing and evaluating ICBMreentry vehicles and penetration aids, in forecastinginstrumentation needs, in monitoring test opera-tions, and in evaluating the test results; (2) researchand development on advanced range instrumentationand associated equipment as indicated by anticipatedneeds; and (3) exploratory research on ICBM pen-etration problems for the purpose of generating andtesting ideas that might contribute to advanced con-cepts and techniques.

Initial work on the first element of the programwas concentrated on assisting the Air Force in theprocurement of the MARS ships and an assessment ofthe other instrumentation on the Atlantic MissileRange. The initial view was that the L-band, C-band,and X-band radars of the first two ships would nothave adequate sensitivity for measurements on low-cross-section reentry vehicles, and that a third shipwith advanced capabilities would be required. Lin-



coln Laboratory recommended changes that wouldincrease the initial capability of the radars and pavethe way for future system upgrades.

Under the second element, Lincoln Laboratoryinitiated a program at Litton Industries to develop ahigh-power L-band klystron amplifier while a planwas being developed for other transmitter research.Sidelobe suppression, the coherence of compressedwaveforms, and the design of burst waveforms toachieve good range and Doppler resolution were sig-nal-design issues of immediate concern.

In slightly more than one year, the ships becameAdvanced Range Instrumentation Ships (ARIS), theprogram became Ballistic Missile Reentry Systems(BMRS), and the ARPA and Air Force technologydevelopment efforts were well integrated. A burst-waveform generator—to be utilized in the ARPAMeasurements Radar (AMRAD) and the TRADEXradar—was developed under the BMRS program.

With limitations on antenna size due to shipboardinstallations, high power was even more important tothe Air Force than it was to ARPA. Under the BMRSprogram, the Laboratory sponsored tube develop-ment programs at Varian, Litton, and Westinghousefor later application in the measurement radars. Lin-coln Laboratory’s Building V, then known as the Ra-

dar Transmitter Research Facility (RTRF), was con-structed to investigate ways to build transmitters withgreater waveform flexibility and to provide modula-tors, power supplies, and cooling with which to testthe new tubes to higher power levels than the vendorscould achieve.

The Laboratory designed a hard-tube modulator,shown in Figure 8, which when operated with eighttubes (six tubes are shown in the photo) could pro-vide pulses of up to 180-MW peak power and 2-MWaverage power. It became the prototype for the AL-TAIR transmitter and was used to test UHF tubes forARIS and ALTAIR and at a later date the tubes thatwould be used for the L & S-band modifications ofthe TRADEX radar. The UHF klystron shown in Fig-ure 9 was destined for use in the advanced ARIS shipsand ALTAIR, but it never achieved its design goals of60-MW peak power and 300-kW average power. As aresult, ALTAIR used a version of the VA-812 klystrondeveloped for the Boston Hill radar, while the ARISsystems used a transmitter with twenty-four travelingwave tubes operating in parallel. Some twenty yearslater the ARIS ships were retired from service, andVarian announced that it would no longer maintainthe facilities needed to build the large VA-812klystron. ALTAIR acquired the ARIS traveling-wave-tube transmitters and is using them to this day.

The RTRF was also equipped with a very high-

FIGURE 8. The hard-tube modulator at the Radar Transmit-ter Research Facility (RTRF) at Lincoln Laboratory. Operat-ing singly, the Westinghouse triode switch tubes could gen-erate 50-kV, 500-A pulses with rise times of 0.5 µsec.Derating was necessary when they were operated in parallel.A pulse transformer stepped the voltage up to the 300 kV re-quired by the klystron.

FIGURE 9. A UHF klystron designed for horizontal operationat 60-MW peak power is shown in the RTRF test stand. Theheavy concrete walls provided necessary shielding againstthe intense X-rays generated by these tubes.



voltage power supply of more modest average power,as shown in Figure 10, in order to test tubes withmodulating anodes.

Project PRESS

Project PRESS (Pacific Range Electromagnetic Signa-ture Studies) was ARPA’s most ambitious DE-FENDER program. It was to include the PINCUSH-ION S-band radar, the TRADEX radar, airborneoptical sensors installed in a KC-135 and Navy A3Daircraft, and numerous ground-based optical sensors.All the sensors were to be interconnected and con-trolled by a central IBM 7094 computer.

In February 1959 the Army decided to site its NikeZeus antiballistic missile system at the KwajaleinAtoll, where it could be tested against missile targetslaunched from Johnston Island or Vandenberg AirForce Base in California. The Zeus system, designedby Bell Laboratories, consisted of the UHF Zeus Ac-quisition Radar, the L-band Discrimination Radar,and the C-band Target Tracking Radar, all located onKwajalein Island, the southernmost island in the

atoll. In order to take advantage of the Zeus targetsfor reentry measurements, ARPA decided to locate itsinstruments there as well. A site on Roi-Namur, thenorthernmost island, was selected for the TRADEXradar by Lincoln Laboratory’s Glen F. Pippert in early1960 [5]. From the outset, ARPA expected the instal-lation to be a research site manned by scientists andengineers. Figure 11 shows Kwajalein Atoll viewedfrom the north. The island in the foreground is Roi-Namur, where all the KREMS radars are located [6].

As the result of a study requested by ARPA andconducted by the Laboratory in 1960, work on thePINCUSHION S-band radar ceased and LincolnLaboratory was given responsibility for the PRESSprogram.

TRADEX

The TRADEX radar, already under construction byRCA when Lincoln Laboratory assumed responsibil-ity for PRESS, was a derivative of the UHF trackersRCA had built for the Ballistic Missile Early WarningSystem (BMEWS), but with an L-band range-track-

FIGURE 10. A 350-kV, 1-A power supply used to test tubes with modulating anodes at theRTRF.



ing and data-gathering capability. With the aid ofpersonnel from the Laboratory and RCA, the radarwas installed and checked out in time to successfullytrack the first Atlas ICBM launched to Kwajalein on26 June 1962. Figure 12 shows the TRADEX an-tenna on Roi-Namur.

TRADEX was one of the earlier radars to use pulsecompression, utilizing a 50-µsec, 1-MHz linear fre-quency modulation, or “chirped,” transmit pulse toachieve high sensitivity, while achieving a range reso-lution of approximately two hundred meters. Withan eighty-four-foot antenna and 2-MW peak power,the TRADEX L-band system achieved a single-pulsesignal-to-noise ratio of 28 dB on a one-square-metertarget at a range of a thousand kilometers, easilyenough to detect warheads as they came over theEarth’s horizon in the vicinity of Hawaii. From thebeginning, the system was coherent and featured anunusually broad range of pulse-repetition frequencies(PRF). High PRFs provided excellent Doppler reso-

FIGURE 12. The TRADEX antenna, located on the island ofRoi-Namur in the Kwajalein Atoll.

FIGURE 11. A view of the northern tip of Kwajalein Atoll looking south. The island in the foreground is Roi-Namur, the site of all the KREMS radars, which can be seen at the extreme left edge of the photograph. Thebrown areas in the picture are the coral reef, exposed at low tide but under three to six feet of water at high tide.



lution for wake measurements, while low PRFs wereneeded to view very long target complexes. As in-stalled, the radar used all-analog circuitry, but it wasmodified over time to provide ever-increasing use ofthe IBM 7094 computer for control, recording, andtrajectory extrapolation. A memorable feature of theearly TRADEX radar was the intermediate-frequency(IF) tape recorder, an analog recorder with three-foot-diameter reels. This tape recorder achieved sufficientbandwidth to record the IF signals and collect the vi-tal phase information, but it did so with very hightape speeds that resulted in spectacular displayswhenever the tape broke.

When the Laboratory assumed control of theTRADEX radar, the first change was to add a pulse-burst waveform to provide improved range and Dop-pler measurements. The burst subpulses were 2-µsecchirps of 20-MHz bandwidth, providing range reso-lution of approximately fifteen meters and allowingthe analysts to examine the amplitude and velocityspectrum of the wake as a function of distance behindthe body. The subpulse spacing of 14 or 28 µsec pro-vided quite high Doppler ambiguities, while the 32-pulse duration of the burst provided good Dopplerresolution. Major modifications to the transmitter topass this waveform and a digital recording system thatcould handle its high data rates were required. Theywere designed and built at Lincoln Laboratory.

During the 1960s, with the developers of both bal-listic missiles and ballistic-missile-defense systems fir-ing missiles into the Kwajalein Missile Range,TRADEX became the primary source of data for re-entry discrimination work. At a time when mostpeople still thought of radar only as a means of mea-suring range, azimuth, and elevation angle, seeingwhat TRADEX could measure on their vehicleschanged the thinking of many of the users.

TRADEX L & S-Band Modification

In 1970 the TRADEX radar was shut down for a ma-jor redesign. The UHF capability was removed, a newfeed and additional channels were added to make itan L-band tracker, and an all-new S-band radar wasadded. The Missile Site Radar of the Nike-X ballistic-missile-defense system (which later became theSAFEGUARD system) operated at S-band, and an S-

band database was needed in order to design discrimi-nation techniques for the system.

Two new transmitter output tubes were developedby Varian for TRADEX and tested at Lexington. TheL-band klystron operated very well at 4-MW peakpower and 300-kW average. The S-band klystron, aderivative of the Missile Site Radar tube, had 250-MHz bandwidth and operated successfully at 4 MWand 120 kW. The tube is so conservatively designedthat three of the four original tubes are still in use.

All of the receiver, signal processing, and recordingelectronics were new. With a Sigma V computer, apowerful real-time computer for the time, the systemachieved a remarkable level of flexibility. The S-bandradar started with only two basic pulses, a 3-µsec, 60-MHz chirp, and a 9-µsec, 17.6-MHz chirp. The sys-tem design allowed the pulses to be used in manycombinations, including trains, bursts, pairs, longpulse bursts, and frequency-jumped bursts, whichprovide a range resolution of one meter. The pulsescheduler allowed mixing many of these waveforms ina single 0.1-sec interval and changing the mix every0.1 sec. The redesigned TRADEX system remainedthe workhorse for the development of discriminationtechniques. In the 1970s everyone in the ballistic-missile-defense (BMD) community had a favoritediscrimination waveform, and TRADEX tried themall.

The TRADEX system to date has covered 545ICBM missions, approximately a hundred experi-ments with interceptors and other locally launchedvehicles, innumerable satellite tracks, orbital-debrisdata-gathering missions, and a host of other experi-ments, such as ionospheric-effects measurements, sea-clutter characterization, and tectonic-plate move-ment. Clearly, ARPA’s vision of a continuouslydeveloping and improving research facility has beenfully realized.

Measurements at White Sands Missile Range

At the same time the TRADEX system was being de-veloped, two other major efforts started. ARPA wasconsidering a next-generation terminal-defense sys-tem and a radar that would serve as a technologytestbed for the system. The system was the ARPA Ter-minal (ARPAT) and the radar was called the ARPA



Measurements Radar, or AMRAD. Lincoln Labora-tory performed several studies of the ARPAT systemfor ARPA under the Radar Techniques Study pro-gram. As a result ARPA requested the Laboratory toassume responsibility for the ARPAT program, andthe Radar Discrimination Technology (RDT) pro-gram emerged. Its goal was (1) to determine the feasi-bility of performing designation and discriminationby using the radar-observed reentry phenomenologyfrom the Reentry Physics program and the PRESSprogram, (2) to study conceptual defense systems us-ing radar sensors, and (3) to build a prototype termi-nal-defense system.

Special Test Vehicle Program

As experiments for the BMRS program were plannedit quickly became apparent that the Atlantic Oceanand Pacific Ocean ICBM tests would not begin toprovide enough payload space for all the plannedmeasurement programs, and that the cost of perform-ing all the tests with ICBM vehicles would be pro-hibitive. The Air Force determined that a vehiclesimilar to the Trailblazer could be used as an eco-nomical means of flying test targets at ICBM veloci-ties. After examination of the instrumentation, range-safety issues, and security at Wallops Island, Virginia,Eglin Air Force Base, Florida, Point Mugu, Califor-nia, and White Sands Missile Range, New Mexico,White Sands was selected as the site for the BMRS ex-perimental program. The security of the inland loca-tion and the prospect of using Nike Zeus test targetsand obtaining data from the Zeus prototype radars atWhite Sands were major factors in the decision.

The Special Test Vehicle program was formed toperform the BMRS experiments. It included thelaunch vehicles and new instrumentation at WhiteSands. The tests were performed with a mix of sound-ing rockets, which could be fired within the bound-aries of the White Sands Missile Range, and with thefour-stage Athena rocket, which like Trailblazer IIused two stages to boost the vehicle to altitude andthe remaining two stages to accelerate the test targetinto reentry at ICBM velocities. Unlike the Trail-blazer, Athena could not be flown on an “out andback” trajectory because of the expected impactpoints of the initial stages. It was launched instead

from Green River, Utah, some four hundred milesnorth of White Sands, and overflew privately ownedland. The sounding rockets achieved a reentry veloc-ity of 4000 ft/sec with a fifty-pound to seventy-pound payload, while the Athena rocket achieved avelocity of 22,000 ft/sec with a nominal seventy-pound payload. The Athena could carry vehicles aslarge as 26 inches in diameter and 72 inches long.The initial plan called for twenty sounding rocketflights and eighty-two Athena launches. To performmeasurements on the vehicles required a large up-grade of the White Sands instrumentation. A sophis-ticated S-band radar system, named RAMPART, de-rived from the earlier PINCUSHION system, and aUHF and L-band radar system called RAM were in-stalled near the impact area on the southern end ofthe range. A similar UHF/L-band system that in-cluded a VHF measurement capability was installedat the more northerly Stallion site to provide high-as-pect-angle data. Because the Special Test Vehicle pro-gram offered ARPA an excellent chance to testAMRAD and its discrimination techniques, WhiteSands was selected as the site for that radar.

AMRAD

The ARPA Measurements Radar, or AMRAD, wasbuilt by the Raytheon Corporation to Lincoln Labo-ratory specifications. An L-band system with a sixty-foot dish, it exploited technologies different fromthose of TRADEX in many areas with the expecta-tion that each radar would be upgraded with success-ful features of the other. While TRADEX experi-mented with pulse compression, AMRAD used burstwaveforms to produce the high Doppler ambiguitiesneeded for wake-velocity measurements. The initialwaveform consisted of a precursor pulse of 0.1 to 8.0µsec in duration followed by up to thirty 0.2-µsecsubpulses with spacings of ten to fifty µsec. To gener-ate the fast rise time needed, the transmitter used anew device, a klystron with a magnetron injectiongun and a modulating anode with a gain of four. Byreconfiguring the waveguide, the system could trans-mit linear or circular polarization, but received onlyorthogonal linear polarizations. The receiver pream-plifiers were Lincoln Laboratory–built cooled var-actor-diode parametric amplifiers, which provided a



system noise temperature of 180 K. Work on the ra-dar began in 1961, and it was turned over to LincolnLaboratory in December 1963 [7].

The AMRAD contract did not include any record-ing capability. Raytheon, under a separate contractfrom Lincoln Laboratory, designed and built a digitalsignal processor for AMRAD. In addition to per-forming all the digital sampling, tracking, and record-ing for the radar, the processor served as the synchro-nizer, providing clocks and frequency sources forother subsystems and all the RF and video start andstop pulses. Logarithmic amplifiers and quadraturephase detectors were used with 6-bit A/D converters.To provide 10-MHz sampling with 4-MHz analog-to-digital converters, three of these units sequentiallysampled the video signal. A phase-shifted 10-MHzclock provided a sampling granularity of 1.5 nsec.The system recorded amplitude and phase samples onthe orthogonal-sum channels and angle-error chan-

nels with up to twenty-one samples on each burstsubpulse. It also collected the metric data, mode in-formation, and samples of the transmitted phase. Allof this was accomplished without a computer! Thissystem represented a great advance from the primitiverecording system used on the Arbuckle Neck radars,and it became the model for systems later installed atTRADEX and ALTAIR.

A highly visible feature of AMRAD was its clutterfence. Detecting very-low-cross-section wakes at theAMRAD site, which was surrounded by mountainsof up to 8000-ft elevation, proved impossible withoutthe addition of a clutter fence. The required fence,designed by John Ruze of Lincoln Laboratory, was104 ft tall and triangular in shape with a perimeter of2000 ft. The clutter suppression it provided matchedtheoretical predictions exactly. Figure 13 shows aphotograph of AMRAD and its clutter fence, “thebiggest corral in New Mexico” [8].

FIGURE 13. The ARPA Measurements Radar (AMRAD) with its clutter fence at the White Sands MissileRange, New Mexico (U. S. Army photograph). In this picture the 60-ft-diameter AMRAD antenna is aimedthrough a gap in the 104-ft-tall fence in order to work with radio-frequency (RF) calibration instrumentationon a boresight tower on a hill a few kilometers away. In actual missile-test operations the antenna is rotatedabout 180° in azimuth and raised in elevation angle, its beam pointing over the fence at the incoming targetcomplex. The fence keeps the antenna’s sidelobes from illuminating the mountain ranges that border theTularosa Valley, where AMRAD is placed, on its east and west sides. Before the fence was installed, strongclutter echoes from these mountains in the same range interval as the target complex made the collection ofuseful data difficult.



Both TRADEX and AMRAD evolved through the1960s; researchers at each system frequently appliedtechniques and hardware developed for the other sys-tem. Pulse compression was added at AMRAD andbursts at TRADEX. RCA-furnished analog recordingwas installed at AMRAD, while a digital system wasinstalled at TRADEX. When it was discovered thatthe wake velocities decayed much more rapidly thanexpected behind the vehicles, the burst parameterswere modified to provide improved Doppler resolu-tion at the expense of lower Doppler ambiguities.

By 1963 the PRESS, BMRS, and Radar Discrimi-nation Technology programs employed 246 membersof the Laboratory technical staff. Laboratory manage-ment started looking for ways to decrease the BMDinvolvement in order to support other areas of re-search. Turning the Arbuckle Neck radars over toNASA was the first step. As a second step, the Labora-tory asked ARPA to be relieved from operatingAMRAD. On 17 February 1966 the responsibilityfor AMRAD was transferred to the Columbia Uni-versity Electronics Research Laboratory, later to be-come the Riverside Research Institute.

ALTAIR

In the early 1960s, the United States was surprised tofind that the Soviet Union was developing very largeVHF and UHF radars (called Henhouse and Dog-house), presumably intended for ballistic missile de-fense and space surveillance. In order to understandhow U.S. strategic weapons would fare against suchradars, it was necessary to test these weapon systemsagainst radars of similar capability. The secondProject PRESS radar was initiated, the ARPA LongRange Tracking and Instrumentation Radar (AL-TAIR). It was specified to track at VHF and collectUHF data with greater sensitivity and spatial resolu-tion than that provided by the TRADEX radar. Avery large antenna and high-power transmitters werespecified to achieve high sensitivity. As a consequenceof these design factors, ALTAIR routinely acquirestargets launched from Vandenberg Air Force Base inCalifornia as they come over the horizon at a range of3500 kilometers.

The ALTAIR antenna is unusual for its size andagility. As shown in Figure 14, the 150-ft diameter

antenna rotates on a 110-ft diameter circular track.To achieve the rates and accelerations necessary totrack reentry vehicles, the antenna is far stiffer thanmost 150-ft dishes. The rotating portion weighs al-most a million pounds. Perhaps this fact was not real-ized at the time, but the system was designed with arecord-setting load on the wheels and track. With ex-tremely hard steel and meticulous alignment, AL-TAIR operated successfully for many years with load-ing more than ten times as high as the worst-caseloads used by the railroad industry. Redesigned bogieswith twice the number of wheels have since been in-stalled to insure long life even with the twenty-four-hour-per-day utilization of the antenna needed forSPACETRACK satellite-tracking operations. To thisday, seeing the antenna operating at its full 2°/sec2 ac-celeration and 10°/sec velocity is impressive.

Transmitter powers of 10 MW peak and 120 kWaverage at VHF and 20 MW peak and 120 kW aver-age at UHF were specified and demonstrated. Origi-nally, three different transmit pulse lengths were pro-

FIGURE 14. The ARPA Long Range Tracking and Instru-mentation Radar (ALTAIR) on the island of Roi-Namur,Kwajalein Atoll, Marshall Islands.



vided at each wavelength to furnish pulse-repetitionrates as high as 3 kHz. The bandwidth of all threewaveforms was the same, providing range resolutionof thirty meters at VHF and fifteen meters at UHF.Control of the system was performed by a HoneywellDDP-224 computer that had only sixteen kilobytesof core memory!

The study to determine the radar specificationswas completed in early 1965. The competitive con-tract to build ALTAIR was awarded to Sylvania. In-stallation and checkout of the system occurred in thelate 1960s and it became operational in May 1970. Inorder to obtain some VHF data at an earlier date, re-searchers modified the TRADEX radar to add a 60-MHz radar and then a 149-MHz capability, whichwas dubbed “SMALLTAIR.”

The ALTAIR contract, like that for AMRAD, didnot include a signature-data-recording capability. TheALTAIR recording system was built by Lincoln Labo-ratory and interfaced to the radar at its IF output.With its relatively broad beamwidth of 3° at VHF,ALTAIR illuminates an entire ICBM complex from afew degrees below the horizon until well into reentry.The recording system provided the ability to range-track and collect extensive signature data on up tofourteen targets at each frequency. Because the datarate was far too great for computer tape drives of thetime, multiple fourteen-channel instrumentation re-corders were used for data recording. After a mission,the tapes were laboriously played back with an 8:1slowdown to transcribe the data to computer tapes forfurther processing. By the late 1960s, when the sys-tem was designed, sidelobes of more than 30 dBdown were readily achievable for the 17.6-MHz-bandwidth waveforms with analog pulse-compres-sion equipment, and reliable 7-bit 10-MHz analog-to-digital converters were available. Control,tracking, and auxiliary data recording were performedby a pair of DDP 224 computers.

SIMPAR

The large UHF search radar of the SAFEGUARDsystem, the Perimeter Acquisition Radar (PAR) wasnot installed or tested at Kwajalein. The PAR softwarecontained a large body of target acquisition, tracking,and impact-prediction software that could not be

tested against realistic targets at the radar’s North Da-kota location. The Simulation of PAR (SIMPAR)program was initiated to test the software. The AL-TAIR UHF system was modified to produce PAR-like data, and the PAR real-time program was run ona Control Data Corporation CDC 6600 computerinstalled at Kwajalein.

The ALTAIR modifications installed in 1973 wereextensive. A new Cassegrainian feed, microwave sys-tem, receivers, and pulse-compression channels wereadded to provide angle tracking at UHF. A frequency-selective subreflector, twenty-two feet in diameter,was developed and installed at the focal point to allowthe system to angle-track at either UHF or VHF. The“coarse or fine” angle-tracking capability thus pro-vided is unique and has been important for many ofthe missions ALTAIR has been asked to perform overthe years. Figure 15 shows the ALTAIR feed.

FIGURE 15. The VHF and UHF feeds of the ALTAIR radar.The white “teacup” (notice its size relative to that of the manstanding below it) is the cover of the conventional five-hornfocal-point feed of the VHF system. The “saucer” is thedichroic secondary reflector of the UHF Cassegrainian sys-tem with a multimode horn at the vertex. The reflector iscomposed of two layers of crossed dipoles that are reso-nant at UHF, making it an excellent reflector at UHF and al-most transparent at VHF. The system angle-tracks at bothfrequencies, and either frequency can drive the antenna-servo system. (A similar frequency-selective subreflectorwas installed on the Millstone Hill radar; see the article byStone and Banner in this issue.)



The PAR waveforms proved difficult to simulate.To operate at PAR pulse-repetition rates, the rela-tively low-duty-cycle ALTAIR transmitter couldtransmit only a 40-µsec expanded pulse. With the re-quired chirp bandwidths of only 115 kHz and 690kHz, the time-bandwidth product of the pulses wastoo low to be compressed efficiently by the analogtechniques of the early 1970s. A digital tapped-delay-line pulse-compression system was designed and builtat Lincoln Laboratory to solve this problem. This sys-tem provided excellent sidelobe performance. Thesystem operated well, and the PAR software producedgood results. The modifications installed were invalu-able for the next ALTAIR modification.

SPACETRACK Modifications

In 1977, U.S. Space Command began to consider anetwork of radars in the Pacific Ocean to detect andtrack new Russian and Chinese satellite launches ontheir initial revolutions. Lincoln Laboratory proposedthat ALTAIR, with its large power-aperture product,could do an excellent job. U.S. Space Command,however, was unsure of both the surveillance scan thatwas proposed and the reliability of the system withthe heavy usage expected. A trial period was arranged,and with heroic effort and good use of the SIMPARsoftware and hardware, ALTAIR began operations byNovember 1977 with its unique 75° bow-tie scan.During the three-month test period, ALTAIR tracked6121 objects out of 6396 assigned, and was far moresuccessful at detecting and tracking newly launchedobjects than the Air Force thought possible.

Because of these excellent results, ALTAIR wasmodified to provide support to U.S. Space Com-mand for both deep-space tracking and the detectionof new foreign launches. New computers, waveforms,and signal processing techniques were required as partof these modifications. Details on the modificationsand ALTAIR’s continuing role as a space-surveillancesensor are found in the previously mentioned articleby Stone and Banner in this issue.

ALTAIR continues to be the most heavily utilizedradar at the Kwajalein Missile Range. During the 128hours per week that ALTAIR works for U.S. SpaceCommand it performs more than 35,000 deep-spacetracks per year and 2500 high-priority near-earth

tracks. For ICBM and interceptor missions ALTAIRis Kwajalein’s primary long-range search and acquisi-tion sensor. Its broad beam and high sensitivity allowit to detect all the targets in a target complex, and itslong wavelength facilitates the identification of thesignificant objects in the complex.

ALCOR

In the late 1960s Lincoln Laboratory analysts becameinterested in wideband radar waveforms for ballisticmissile discrimination. The rationale for this interestis explained in more detail in the article entitled“Wideband Radar for Ballistic Missile Defense andRange-Doppler Imaging of Satellites,” by William W.Camp et al., in this issue. In short, wide-bandwidthradars provide a means of measuring the length ofwarheads and decoys. With short pulses on a staticrange, length was easily measured, but the effects ofthe reentry plasma sheath on the measurement capa-bility were difficult to predict. In addition, the gen-eration and processing of wideband waveforms withsufficient energy to make a practical BMD radar wasa challenging task. In response to a Lincoln Labora-tory proposal in mid-1965, ARPA authorized theLaboratory to build the ARPA Lincoln C-band Ob-servables Radar (ALCOR). Acting as the prime con-tractor, Lincoln Laboratory built the radar with assis-tance from RCA, Westinghouse, Hughes, Honeywell,and numerous smaller contractors. ALCOR becameoperational in January 1970. Figure 16 shows the an-tenna and radome during installation of ALCOR onRoi-Namur.

The original waveforms were 10-µsec chirpedpulses of 6 MHz and 512-MHz bandwidth operatingat a peak power of 4 MW. The key to processing the500-MHz waveforms was stretch processing or time-bandwidth exchange. The stretch-processing tech-nique and additional details about ALCOR’s wide-band performance are discussed in the previouslymentioned article on wideband radar by Camp et al.in this issue. An unusual feature of ALCOR, whichreduced the system cost and insured the match be-tween all receiver channels, was to multiplex all thesignals with delay lines and pass them sequentiallythrough a single set of signal processing hardware.Many of the components, including tapped-delay-



line compression networks, analog-to-digital convert-ers, and the control computer, were the same as thoseused on ALTAIR.

With its beamwidth of only one-third of a degreeand limited range window, ALCOR needed very ac-curate pointing information to acquire targets. AL-TAIR and TRADEX were excellent sources of suchinformation. The role of the PRESS Control Centerand its IBM 7094 computer was expanded to accom-modate the new radars. Files from all three radars aswell as external sources were stored, smoothed, ex-trapolated, and redistributed to the radars as neededto provide designation. Another ALCOR first was theuse of surface-acoustic-wave devices built by theLaboratory to provide all-range compression of the500-MHz pulses. (See the article entitled “Radar Sig-nal Processing,” by Robert J. Purdy et al., in this is-sue.) Surface-acoustic-wave technology found wideapplication for analog pulse compression and waslater used at TRADEX and ALTAIR.

ALCOR was the nation’s first high-power, long-range wideband radar. It played a pivotal role in intro-

ducing the BMD community to the discriminationpotential of wideband systems. As a consequence,wideband capability is a key feature in today’s BMDsystems. ALCOR’s impact on space-object identifica-tion has been equally profound. Early experimentswith ALCOR led to the nation’s impressive capabili-ties to create high-resolution images of satellites atgreat distances.

The Formation of KREMS

In the late 1960s, as offensive-weapon testing in-creased at the Kwajalein Missile Range, the Directorof Defense Research and Engineering, John Foster,insisted that the sensors at Kwajalein, includingARPA’s PRESS, be treated as a national facility. Hedirected that the PRESS complex be transferred tothe U.S. Army, and that the sensors be kept abreastof, or lead, advances in state-of-the-art radar technol-ogy. The Army submitted a support plan that in-cluded the required assurances. In 1969 TRADEXand the PRESS complex were transferred to the U.S.Army; ALTAIR and ALCOR followed as soon as theywere declared operational. In a formal ceremony atthe Pentagon, the Kwajalein radar complex was re-named the Kiernan Reentry Measurement Site(KREMS) in honor of Lt. Col. Joseph M. Kiernan,U.S. Army, a visionary ARPA PRESS program man-ager who died in action in Vietnam. Kiernan man-aged the program from 1963 until 1966 and led thestudies that resulted in ALTAIR and ALCOR.

Real-Time Experiments

Using recorded data, analysts had shown that ballisticmissile discrimination during reentry was possible,but there was concern in the BMD community aboutthe practicability of the solutions being proposed.Could the job be done without human intervention,and were the computers powerful enough to performdiscrimination in real time? To answer these ques-tions, Lincoln Laboratory proposed the Reentry Des-ignation and Discrimination (REDD) system. Usinga CDC 6600 computer, the Laboratory developedreal-time discrimination algorithms, incorporatedthem into a complete discrimination logic (schema),and tested them on preprocessed recorded data. Theschema was then installed in a second system at Roi-

FIGURE 16. The ARPA Lincoln C-band Observables Radar(ALCOR) antenna and radome during installation on Roi-Namur. At C-band frequencies the forty-foot antenna pro-vides a beamwidth of one-third of a degree, which is excel-lent for sensitivity and angular accuracy but difficult to pointaccurately.



Namur for true real-time tests on ICBM targets. Thesecond system included a CDC 6600 and special-purpose hardware that tapped into the data fromTRADEX and ALTAIR and performed the prepro-cessing that had been done in software at Lexington.The REDD system was used in many ICBM tests atKwajalein with good success, and it alleviated manyof the concerns about real-time discrimination.

Control Center Development

The initial goal of the PRESS Control Center was toassist the narrowbeam optical sensors and theALCOR radar in acquiring targets by providing di-recting data based on stored target trajectories andsatellite ephemerides, up-range track data, andsmoothed and extrapolated track data fromTRADEX and ALTAIR. In operation, it quickly be-came apparent that there were other important ad-vantages in operating an integrated system. With datafrom all the sensors, non-nominal target deploymentsand sensor problems and mistakes were more appar-ent to the PRESS Control Center operators than theyhad been to the sensor operators. PRESS ControlCenter personnel were able to direct corrective actionthat saved valuable data on the one-of-a-kind mis-sions typically performed at Kwajalein.

Working together and using each other’s tracks fordesignation, ALTAIR and TRADEX could stepthrough the target complex and identify the impor-tant objects far more rapidly than when they tried toperform the job independently. ALCOR beacontracks that provided positive identification of a fewobjects in a target complex proved valuable in under-standing the deployment of the whole complex. Amission goal of determining how UHF chaff de-ployed around the dispensing vehicle was easily metby beacon-tracking the dispenser with ALCOR andslaving ALTAIR’s recording window to the ALCORtrack. When a millimeter-wave radar was added tothe system at a later date, its ability to provide real-time images became an important tool to help in un-derstanding the target-complex deployment.

When the site was named KREMS, the PRESSControl Center was renamed the KREMS ControlCenter (KCC). As the KCC role grew, the IBM 7094was replaced first with the CDC 6600 and then with

multiprocessor Digital Equipment Corporation Al-pha computers. Changes included the application ofdetailed sensor-bias models and logic to determinethe best source of directing data for each object in thecomplex. Metric data to and from the sensors and sta-tus and cross-section data from the sensors to theKCC were passed via an Ethernet network connec-tion at an update rate of 20 Hz.

As the power of computer-generated graphics in-creased, a number of site-wide displays were also de-signed and the KCC provided the data to update thedisplays at the sensors via a second Ethernet networkconnection.

In 1990 the U.S. Army requested that LincolnLaboratory assist them in integrating activities at theKwajalein Missile Range by designing a control centerfor both KREMS and the other sensors and systemsof the Range. The Kwajalein Mission Control Center(KMCC) was the result. It is located on the main is-land of Kwajalein and provides the capability tomonitor the status of the mission, weather, and allKwajalein sensors, and to control all the radars, opti-cal systems, and telemetry assets of the KwajaleinMissile Range.

Millimeter-Wave Radar

Discussions between Lincoln Laboratory and theArmy’s Ballistic Missile Defense Advanced Technol-ogy Center (BMDATC) in 1977 led to a decision tobuild a dual-frequency millimeter-wave radar operat-ing at 35 GHz and 95.5 GHz with 1000-MHz band-width waveforms. Lincoln Laboratory was the primecontractor for the development of this radar. The ra-dar had a variety of goals and offered many chal-lenges. BMDATC was interested in using millimeter-wave seekers in interceptors and wanted to developcomponents for operation at the short wavelengths inaddition to collecting a signature database on ballisticmissiles. BMDATC and SOI (Space Object Identifi-cation) personnel were also interested in the discrimi-nation and imaging possibilities of the short wave-length and 0.25-m range resolution.

A major challenge in the development of millime-ter-wave radar was the generation and transmission ofmicrowave power. The selected peak powers of 30kW and 6 kW at the two frequencies resulted in ex-



tremely high power densities in the very smallwaveguide and RF structures used at these frequen-cies. Varian, the transmitter-tube developer, wentthrough many iterations of geometry, materials, andprocessing before producing guns and focusing struc-tures that could provide the current density needed.Particularly at 95 GHz, the high loss of RF power perunit length of the waveguide presented problems.Combining the output of two tubes proved impracti-cal because the loss of the combining network almostequaled the benefit of the second tube. Even with thetransmitters and receivers mounted on the antenna asclose to the feed as possible, the RF losses were high.

A second challenge was the design of the antenna.With the limited power of the system, a large 13.7-mantenna was required to achieve the required sensitiv-ity. The antenna was designed to be an extremelyrigid structure, and great attention was placed onmaintaining it at a uniform temperature, in order toachieve adequate surface tolerances on such a largedish. The antenna beamwidths (0.042° and 0.014°)are very small compared to the ALCOR beam, whichhad proven difficult to point. With the advances intrajectory-extrapolation algorithms and calibrationthat have taken place, the millimeter-wave (MMW)radar acquires targets very reliably and achieves angleaccuracy typical of a good optical telescope, approxi-mately 40 µrad. Figure 17 is a photograph of theMMW radar antenna, taken as the radome was beinginstalled.

As expected, the MMW radar has become the pre-mier imaging radar of the Kwajalein Missile Rangeand is heavily used by the SOI community. Its shortwavelength effectively increases the number of scat-terers visible on objects being observed and contrib-utes to the superb detail of the images. Its ability tomeasure miss distance or impact point on intercepts isincreasingly important to the BMD users. Additionaldetails on the MMW radar and its uses appear in thepreviously mentioned article on wideband radars byCamp et al. in this issue.

Evolution of KREMS

Ballistic missile defense has always been a rapidlychanging process, driven by changing threats, mis-sions, and technological opportunities for improved

capability. In accordance with ARPA’s and John Fos-ter’s vision, the four KREMS radars have undergone acontinuous process of modification and upgradesince they were constructed. The site manning hastraditionally included a cadre of radar-system designengineers to support the effort to stay at the forefrontof technology.

All of the KREMS radars have had their comput-ers, recording, and display systems upgraded severaltimes and added coherent integration capability to as-sist in acquiring targets at longer ranges.

At TRADEX, long continuous-wave and pseudo-random phase-modulated noise waveforms wereadded at S-band to emulate Soviet BMD radars. Anarrowband 565-µsec pulse was added at L-band forlong-range acquisition and deep-space tracking. A50-µsec, 20-MHz chirp waveform provides improvedrange resolution for complex multitarget missions,and a multitarget tracker provides noncoherent inte-gration, target detection, acquisition, and target iden-tification aids on up to sixty-four targets. Most re-

FIGURE 17. Construction of the MMW (millimeter-wave) ra-dar and radome on Roi-Namur, Marshall Islands. The heavystructure of the dish is needed to achieve the surface toler-ance required for the short wavelengths of the radar.



FIGURE 18. Configuration of the 35-GHz beam-waveguide system.

cently, a new frequency-jumped burst that simulates awaveform being considered for the U.S. Navy Aegissystem has been added.

In addition to the SIMPAR and SPACETRACKmodifications, ALTAIR has seen many other majorchanges. The UHF transmitter replacement, necessi-tated by the vendor’s inability to make replacementsfor the 20-MW, 115-kW klystron, proved so benefi-cial that when the second ARIS traveling-wave-tube(TWT) transmitter became available, ALTAIR ac-quired it to augment the transmitter with an addi-tional eight TWTs to reach an average power of 330kW. For deep-space tracking doubling the averagepower halves the coherent integration time requiredand thus doubles the number of targets the radar cantrack in an interval of time. The TWT transmitter ismore efficient than the old klystron transmitter. Witharound-the-clock operations, the savings on thepower bill are impressive.

To fully utilize the capability of the transmitter,which has a peak power rating of only 6.5 MW, AL-

TAIR needed a number of new, longer UHF pulses. Avery large finite-impulse-response (FIR) digital signalprocessor was implemented to generate and compressthe new waveforms and replace all the existing pulse-compression hardware and software. The 3000-tap-per-channel FIR filter required a “world-record” gate-array chip to be designed. With the new design, eightracks of equipment—one per receiver channel—re-placed twenty-four racks of analog hardware.

The most recent ALTAIR modification providesthe capability to change the VHF transmit polariza-tion in real time so that the full scattering matrix of atarget can be explored. The modification also pro-vides polarization and other waveform changes on apulse-by-pulse basis.

At ALCOR, beacon tracking, pulse pairs, and mul-tiple range windows were added in the 1970s. Withthe additional computer power available in the1990s, ALCOR has doubled its PRF and added a sec-ond independent range tracker.

By the late 1980s, radar signal processing technol-

Circularpolarizer

To subreflector

Frequency-selectivesurface

Virtual images

Multimodetracking feed

OPreceiver

PP∆Az

Two-tubecombiner

High-power path

Comparator

Cassegrainfocalpoint

Polarizationfilters

45° Faradayrotator

Final poweramplifiers

∆El

Receive path

Radiatinghorns



ogy had advanced enough to permit significant per-formance improvement of the MMW radar. Im-proved computers and digital signal processing equip-ment allowed full PRF real-time compression of thewideband pulses and coherent integration to providesignificant improvement in sensitivity. Continuingdevelopment of coupled-cavity TWTs allowed Varianto produce 35-GHz transmitter tubes with 50-kWpeak power and 2-GHz bandwidth.

The application of new quasi-optical techniquesprovided an outstanding improvement in the micro-wave system [9]. By moving the radiating horn closeto the tube and then directing and refocusing thebeam with a series of mirrors, Laboratory researchersdesigned and implemented a beam-waveguide systemwith excellent properties. Microwave losses were re-duced by more than 3 dB, and the power-handlingcapability was greatly increased. Better antenna illu-mination with lower sidelobes, deeper monopulsenulls, better polarization isolation, and bandwidthwell over 2 GHz were all achieved. With transmitter

and microwave systems capable of 2-GHz band-width, the rest of the system is now being modified toprovide 2-GHz waveforms and a second independentrange window, which will provide more precise mea-surements of interceptor hit point or miss distance.

Figure 18 illustrates the configuration of the 35-GHz beam-waveguide system. Figure 19 is a photo-graph of the 95-GHz scalar or ridged horn feeds.These very small horns launch an almost perfectGaussian beam. After refocusing with the mirrors ofthe beam-waveguide system, excellent illumination ofthe 13.7-m MMW radar antenna is achieved. Thecontrast in the size of these feeds and those of AL-TAIR is a reminder of the very broad spectral bandthat is covered by the KREMS radars.

The Future

While new technology has been applied at KREMSto improve capability and replace unsupportable ob-solete equipment, it has not been used specifically toreduce operating costs. The present systems are com-plex and require substantial numbers of highly skilledpersonnel to operate and maintain them. Supportingthose personnel and their families at a remote island istoo costly in a time of shrinking defense budgets. Amajor effort is under way to deal with this challenge.The systems are being modernized, remoted, and au-tomated to reduce their operations and maintenancecosts. Replacement of special-purpose processors andone-of-a-kind electronics with powerful general-pur-pose computers and commercial off-the-shelf digitalhardware will simplify the systems. Coupled withbuilt-in diagnostics to detect and isolate faults to thecircuit-board level, the capabilities of the new tech-nology will greatly reduce the required number andskill level of maintenance personnel. Enforcing acommon design for all the radars will facilitate main-tenance by a matrixed operations and maintenanceorganization, and reduce the implementation costs aswell. Remoting the operations and diagnostics fromRoi-Namur to the main island of Kwajalein will re-duce intra-atoll transportation costs and further fa-cilitate the matrixed support organization. Remotingsoftware development to the continental UnitedStates will allow additional reductions in island per-sonnel. Finally, the systems will be more tightly inte-

FIGURE 19. The scalar, or ridged, horns used for the 95-GHzradar. This small horn launches an almost perfect Gaussianbeam that, with a properly designed optical system, providesexcellent illumination of the 13.7-m primary reflector. Thesystem provides higher power capability, broader band-width, lower losses, and better antenna patterns than a con-ventional waveguide system could at shorter wavelengths.



R E F E R E N C E S1. Reentry Physics Program Semiannual Technical Summary

Report to the Advanced Research Projects Agency, LincolnLaboratory (8 Oct. 1959).

2. W.G. Clay, M. Labitt, and R.E. Slattery, “Measured Transitionfrom Laminar to Turbulent Flow and Subsequent Growth ofTurbulent Wakes,” AIAA J. 3 (5), 1965, pp. 837–841.

3. L.J. Sullivan, “The Early History of Reentry Physics at Lin-coln Laboratory,” Linc. Lab. J. 4 (2), 1991, pp. 113–132.

4. Radar Techniques Study Semiannual Technical SummaryReport to the Advanced Research Projects Agency, LincolnLaboratory (31 Dec. 1959).

5. M.S. Holtcamp, “The History of the Kiernan Reentry Mea-surements Site,” Kwajalein Missile Range Directorate, BallisticMissile Defense Command, Huntsville, Ala. (1 Oct. 1980).

6. K.R. Roth, M.E. Austin, D.J. Frediani, G.H. Knittel, andA.V. Mrstik, “The Kiernan Reentry Measurements System onKwajalein Atoll,” Linc. Lab. J. 2 (2), 1989, pp. 247–276.

7. E.C. Freeman, ed., MIT Lincoln Laboratory: Technology in theNational Interest (Lincoln Laboratory, Lexington, Mass.,1995), p. 83.

8. J. Ruze, F.I. Sheftman, and D.A. Cahlander, “Radar GroundClutter Shields,” Proc. IEEE 54 (9), 1966, pp. 1172–1183.

9. W.D. Fitzgerald, “A 35-GHz Beam Waveguide System for theMillimeter-Wave Radar,” Linc. Lab. J. 5 (2), 1992, pp. 245–272.

grated and automated to reduce the demands on op-erators and increase the capability to handle the com-plex multiple-target-multiple-interceptor missionsexpected in the 21st century. The overall goal of thismodernization program is to make the KwajaleinMissile Range sensors operate as a single multibeam,multispectral sensor with operating costs reduced byas much as 50%.

Summary

Lincoln Laboratory entered the field of ballistic mis-sile defense in the mid-1950s at the inception of theICBM. In the ensuing forty years, Laboratory engi-neers and scientists, aided by an active and innovativeprogram in the development of radar systems andmeasurement techniques, have led the nation in thefield of ballistic missile discrimination. The discrimi-nation techniques and radar advances developed byLincoln Laboratory undergird the missile-defense sys-tems being developed today. The four radars of theKREMS complex, which have contributed mostheavily to our understanding of the field, continue toplay a vital role. With the modifications now beingintroduced, they will continue to do so well into thetwenty-first century.



. is a part-time senior staffmember in the Field Systemsgroup. He received an A.B.degree in physics from Will-iams College in 1953 and anM.S. degree in applied physicsfrom Harvard University in1959. He joined the RadarTransmitter group at LincolnLaboratory in 1962. His firsttask was to build and equipthe Radar Transmitter Re-search Facility and to designand construct a 200-MWhard-tube modulator to test anumber of klystrons beingdeveloped for Lincoln Labora-tory. He monitored the devel-opment of the ALTAIR andALCOR transmitters and wasthe lead Lincoln Laboratoryengineer for the major L & S-band modifications of theTRADEX radar. He went toKwajalein Island in 1970 asthe TRADEX section leader tooversee the installation of theTRADEX modifications. Aftertwo years, he became ALTAIRsection leader for the installa-tion of the SIMPAR modifica-tions and spent one year as theKREMS associate site man-ager. Returning to Lexingtonin 1974, he spent four years asassistant leader and thenassociate leader of the InfraredRadar group, working on theFirepond radar. He joined theField Measurements group andworked on various aspects ofthe Kwajalein Missile Rangeprogram as a staff member,assistant leader, and associateleader. He retired in 1996.

. received an S.B. degree inelectrical engineering fromMIT and an M.S. degree inphysics from the University ofIllinois. He joined LincolnLaboratory in 1952. His earlywork at the Laboratory in-cluded design and program-ming of the intercept functionof the Cape Cod System andof the Experimental SAGESubsector (both prototypesystems of continental airdefense), and analyses of theacquisition and tracking capa-bilities of the Ballistic MissileEarly Warning System radars.From 1963 to 1969 he wasassociate leader and thenleader of the Systems Analysisgroup, where he was engagedin the design, fabrication, andperformance evaluation ofballistic missile penetrationaids. From 1969 to 1993 heserved as assistant head, associ-ate head, and head of theRadar Measurements division.He is now a consultant to thedivision as well as a member ofthe Independent Science andEngineering group that advisesthe Ballistic Missile DefenseOrganization on mattersrelated to ballistic missiledefense. Among his awards arethe Outstanding CivilianService Medal, conferred bythe Secretary of the Army. Heis a lifetime Senior Member ofthe IEEE, a Senior Member ofthe American Institute ofAeronautics and Astronautics,and a member of the Ameri-can Physical Society, theAmerican Association for theAdvancement of Science, andthe Society of Sigma Xi.

Radars for Ballistic Missile Defense Research · • INGWERSEN AND LEMNIOS Radars for Ballistic...

Documents

Transcript of Radars for Ballistic Missile Defense Research · • INGWERSEN AND LEMNIOS Radars for Ballistic...