The measure of MASINT
Journal of Electronic Defense
Copyright UMI Company 1998. All Rights Reserved. Copyright Horizon House Publications, Inc. Aug 1998
Improvements to existing reconaissance payloads, such as the multiband upgrade for the U-2's Senior Year electro-optic sensor, have catapulted these systems into the MASINT arena. (Lockheed Martin photo)
When all's said and done, the US intelligence community is a technical one. For all the romanticization of Cold War spyversus-spy exploits, the fact is the United States broke even, at best, in its human - or, more accurately, HUMINT - campaigns of the past 50 years. The Aldrich Ames case is only the most recent, if not perhaps the most damning, testament.
But the US technical-intelligence system - or system of systems - remains unmatched, a marvel to behold, the embodiment of a scientific golden age when government and defense innovation drove industrial modernization. The US's unquestioned dominance in this forum continues today under the commercially driven information revolution, also a US phenomenon. Technological wizardy is one facet of US ascendancy in the electromagnetic spectrum; another is sheer volume. US technical collectors rove the globe in successive layers, from spaceborne satellites to undersea systems, from antenna-festooned aircraft to the more prosaic man-with-the-cam (in this day and age, often as likely to be recording IR signatures as snapping traditional photos at, say, international arms shows and such venues). These machine-based spies are responsible for most of the product circulating through Central Intelligence Agency (CIA) and Defense Intelligence Agency (DIA) analytical offices. The agencies that acquire, operate, maintain and manage these systems receive by far the lion's share of the total national intelligence budget.
As they usher in the so-called revolution in military affairs, fast-moving information technologies may do more than merely increase the potency of each individual intelligence discipline. The US may also see a revolution in its very concept of technical intelligence. A reshuffling is on tap, bringing new concepts and approaches to the fore, a whole new prioritization and emphasis in intelligence collection and exploitation. It already has a name, just not much name recognition: MASINT.
TAKING THE MEASURE
Technical intelligence consists of three primary disciplines, or "INTs," for collecting information in the electromagnetic spectrum. The first two are well known to JED readers. They have become in fact almost synonymous with technical collection after a storied history of Cold War service: signals intelligence (SIGINT), the exploitation of communications, radar, data and other electronic transmissions; and imagery intelligence (IMINT), the gathering of information through various electro-optic (EO) means, including photographic, radar imaging and infrared (IR) imaging.
The third discipline, for most of its existence the overlooked runt of the litter, is measurements and signatures intelligence (MASINT). But MASINT may soon develop a new and powerful musculature. IC 21: Intelligence Commmunity in the 21st Century, the influential 1996 study by the House of Representatives' Committee on Intelligence, predicts a radically more complex threat environment in the next century, where even technologically unsophisticated states will have access to inexpensive but high-tech weapons. At the same time, effective denial and deception techniques will become pandemic. The US intelligence community (IC). write the authors, will need to employ new capabilities if it is to meet its consumers' needs: "One such capability is Measurement and Signature Intelligence, or MASINT.
MASINT...can provide unique contributions to the IC in terms of specific weapon identification, chemical compositions, material content, etc. Such unique identifications will be a major factor in answering the future questions of `who, what, where, when and why.' In fact, some believe MASINT will be the most important `technical INT` of the future."I According to the official definition, MASINT is "technically derived intelligence that detects, locates, tracks, identifies and describes the specific signatures of fixed and dynamic target sources." If this definition seems aggravatingly vague, it is because MASINT is indeed hard to characterize neatly. It is a sprawling discipline, a passive discipline, overlapping but not conforming to accepted notions of SIGINT, IMINT and sometimes even HUMINT.
MASINT capabilities actually encompass a disparate gaggle of technology disciplines and activities. Officially, it includes radar, laser, optical, IR, acoustic, nuclear radiation, radio frequency (RF), spectroradiometric and seismic sensing systems, not to mention gas, liquid and solid materials sampling and analysis (also known as effluent and debris sampling). An official at the DIA's Central MASINT Office (CMO) made this useful distinction: "The other predominant intelligence areas tend to be defined by their place in the electromagnetic spectrum. MASINT actually covers the entire spectrum and beyond, but can effectively augment SIGINT and IMINT." As described in a 1997 Department of Defense (DOD) Inspector General's report, MASINT systems use SIGINT and IMINT data, but subject that data to specialized processing, using data streams not otherwise tapped.2 MASINT is a "non-literal" discipline. It feeds on a target's unintended emissive byproducts, the "trails," - they spectral, chemical or RF that an object leaves behind. These trails form distinct signatures, which can be exploited as reliable disriminators to characterize specific events or disclose hidden targets. In formal terms, MASINT captures metric data (hence the "measurements" in the MASINT acronym) derived from the direct measurement of the kinematics of the targets of interest (e.g, the flyout pattern of a ballistic missile); signature data are derived from high-fidelity measurements that allow unique identification of the targets.3
Historically, MASINT has been more a catch-all term than a coherent discipline, a name used to describe a series of disjointed, loosely defined intelligence-collection techniques used during the Cold War to gather extremely fine-grained threat data on foreign (primarily Soviet) strategic capabilities. The original MASINT mission, which continues today, was exclusively national intelligence, providing data to support treaty monitoring, proliferation management, technical exploitation and weapons assessment.
MASINT systems' biggest clients have been agencies like the National Air Intelligence Center, which catalogs EO and radar data on foreign missile and aircraft flight characteristics. Perhaps the best-known MASINT asset is the Cobra Ball program, the RC-135 sister aircraft to the Rivet Joint that uses very sensitive, very long-range thermal sensors to capture data on nuclear and ballistic-missile testing.
For disciples of the combatsupport intelligence agencies like the National Security Agency (NSA), MASINT has always been something of a peculiarity, an egghead pursuit that fell not into a true intelligence category but into a "scientific and technical" (S&T) classification. Indeed, the moniker was deserved.
MASINT functioned as an archival discipline, the task of analyzing and exploiting collected data an immense laboratory task that could take weeks -- months even - to turn out a finished intelligence product.
That all changed with the advent of the modern computer processor, which has blurred the lines between tactical and strategic, operational and scientific intelligence collection. In the past, according to the CMO source, collected MASINT data "used to have to go through the three bowels of development down in some building, and go through massive S&T exploitation to determine its signature. That is still necessary. But what has happened is, with some of these systems and the incredible rate of processing speeds today, that kind of process and signature extraction can now be done in nanoseconds instead of weeks. That's literally, in a word, why the change - it's processing."
MASINT will make ground-vehicle identification easier for suppresion-of-enemy-air-defense missions, even if the threat sensors are passive or the radar is not emitting. (Thomson CSF-photo)
The change to which he refers is the aggresive reorientation of MASINT to battlefield support known by the official acronym SMO, for support to military operations over the past several years. The revolution in processing power, microelectronics, data dissemination and miniaturization, accompanied by the changing nature of global conflict, has opened portals to a new world of applications for MASINT. The military and intelligence communities' ardent embrace of SMO has provided the forward impetus. The result: in 1993 - 41 years after the creation of the NSA, the US SIGINT temple -MASINT finally gained its own identity, consecrated in bureaucratic purple with the establishment of the CMO. Located within the DIA, the CMO is ordained as the central functional manager for national and DOD MASINT activities. The essence of its charter is to orchestrate technology development and support to various users. Without a doubt, SMO tops the list. "Intelligence for the future battlefield" was how US Army MG John Leide, the first director of the CMO (now retired), described MASINT upon his appointment.
THE SMO SHOW
MASINT has all the right qualities to appeal to a 21st-century warfighter. It punc hes all the RMA hot buttons: it's multi-INT, cross-disciplinary, all source, non-stovepiped. Battlefield users will soon get a taste of MASINT's S&T capabilities - although they may not know it - with the advent of the new generation of "brilliant" weapons. Laboriously refined, MASINT-derived three- dimensional LADAR data, for example, will give the SAM-, SCUD- and tank-hunting Low-Cost Autonomous Attack System (LOCAAS) the genius to discriminate among all types of ground vehicles, and even to identify targets down to specific makes (see "The A to Z of SEAD," JED, July 1997, p. 35 for more on LOCAAS).
The real question, however, is whether MASINT can serve actual, actionable tactical intelligence needs. Can it feed timely -- meaning realtime - information to the fighter cockpit, the tank driver, the specialforces trooper, the coordinate-hungry precision-guided munitions? Over the past few years, the CMO has launched a number of cooperative developments, demonstrations and even deployments to prove its readiness for military support. According to the CMO source, current systems and near-term efforts have focused on indications and warning; counter-concealment, camouflage and deception (CCD); theater missile defense; battle damage assessment; and chemical/biological warfare. An overview of some current MASINT activities might best be organized by discipline area.
Multi/Hyper/Ultraspectral: The most active MASINT domain today is spectral radiometrics - not surprising, considering its close association with the well-known EO/IR technologies that outnumber all other sources of technical intelligence. Unlike traditional IMINT systems, however, MASINT devices do not capture spatial images but perform feature extraction, scouring targets in the tens (multispectral), hundreds (hyperspectral) and even thousands (ultraspectral) of spectral bands.
Multispectral imaging, the bottom rung of the MASINT ladder, is the most easily attainable. It is perhaps better described as an augmentation to IMINT than as a coherent MASINT capability unto itself. Multispectral systems are entering service today: the Senior Year EO Reconnaissance System, an upgrade to the U-2's baseline camera system performed by Hughes Danbury, incorporates about seven spectral bands through the use of special filters. Cobra Brass is a CMO-sponsored "overhead" effort to provide multispectral reconnaissance for the Space-Based Infrared System. The Cobra Brass sensor uses a staring focal-plane array and a bandswitchin, filter wheel to fastframe through different spectral settings. According to open - source reports, a major thrust of the program is to increase the timeliness of the system's tasking and reporting , allowing the imagery to be processed in real time through the existing theater infrastructure. Uses could include theater missile defense and other battlefield-characterization missions.4 Demonstration efforts have already produced some satellite-based intelligence; launch of an operational Cobra Brass sensor was scheduled for this year.
Next into service will be hyperspectral sensors, which cover a much wider frequency range. For instance, TRW has built a camera that views targets in 384 spectral bands. The most immediate tactical demand for such fine-grained discrimination seems to be counter CCD - in other words, finding and identifying concealed vehicles, missile launchers and other high-value targets. As explained by the CMO source, "in the spectral reality in which we all live, any object emits or has reflectivity or emissivity from the very body that it's wearing." For example, a hyperspectral sensor could differentiate between a pair of dark-blue pants and a light-blue shirt.
Of course, so could a high-resolution color TV camera. The utility of hyperspectral sensing becomes evident where literal imaging fails. A good example is a tank covered with camouflage netting and hidden in trees. The standard EO sensor won't make out the corners or edges needed for a spatial identification. The hyperspectral sensor, however, will read the spectral signatures that sunlight or ambient light is reflecting at the sub-pixel level. Any subspectral anomaly - - metal, paint, the synthetic material of the netting - will be picked out and arrayed in revealing color on the display.
To date, the marriage of hyperspectral cameras with unmanned aerial vehicles (UAVs) has produced the most mature exploration of the technology's tactical potential. In 1996, the CMO conducted a demonstration dubbed Covered Lantern using a small, Naval Research Lab (NRL)-developed hyperspectral sensor mounted on a Pioneer UAV. The aircraft conducted about nine flights, searching for tanks, SAM launchers and other military targets disbursed over a wide swath of farmland and townships, concealed under camouflage and foliage. More Covered Lantern flights will occur during the Air Force's Expeditionary Force Experiment (EFX) '98, slated for next month, and during the All Services Combat Identification Evaluation Team (ASCIET) '99 exercise, which aims to reduce or eliminate fratricidal friendly-fire incidents. The NRL has also embarked on a program called Dark Horse, a four-year effort to develop and demonstrate real-time hyperspectral detection, cueing, target location and target designation for an autonomous uninhabited combat air vehicle. The first demonstration, a daytime event, occurred last November over the Patuxent River Naval Air Station in Maryland. The platform was a P-3, equipped with a high-resolution hyperspectral framing camera (25 million pixels), a pulsed laser (for target designation) and a datalink pod to transmit the imagery to a ground station at NRL headquarters in Washington. According to the NRL, the system successfully detected and cued on both ground targets and on an airborne target aircraft flying 4,000 ft below the P-3. The next round of tests will use an IR hyperspectral sensor for day/night operation.
Ultraspectral performance - coverage from ultraviolet through far IR - is the next frontier. The hardware, the lightweight, low-cost sensor heads, could be available off the shelf within a few years.
Ultraspectral exploitation may range into realms not traditionally associated with imagery. For instance, said the CMO official, an ultraspectral sensor might be employed to "split" the source energy from a rocket motor, identifying and typing the missile by its consituent fuels and materials. Processed in real time onboard an aircraft, this information could feed into a defensive aids suite. There is also intense interest in applying ultraspectral techniques to chemical agent detection, added the official. The CMO's Chemical Detection Remote Sensing System program is investigating dropsonde chemical detectors, ultraspectral sensors that can analyze gas particles for distinctive signatures (e.g., nerve gas), then transmit an alert to nearby soldiers on a chemical/biological warning net.
Acoustic/Seismic: In accordance with its cross-disciplinary agenda, the CMO is sponsoring a number of programs in the acoustic/seismic arena - off the end of the electromagnetic spectrum - also looking at vehicle detection/identification/targeting and chemical-agent detection.
The Steel Rattler, an unattended acoustic/seismic device oriented toward theater missile defense, made an effective debut at the 1997 Roving Sands demonstration. The system is intended to provide discriminating detection and targeting in the pre-launch and pre- deployment stages, cueing strike assets before the missile threats are armed and dangerous. Bearing little resemblance to Vietnamera "bump counters," the Steel Rattler carries an embedded computer that can instantaneously process sound waves produced by moving vehicles. According to the CMO official, the system can make specific identifications, distinguishing a HUMVEE from a tank from a missile launcher, based on its acoustic signature database. The system will also take part in the ASCIET and EFX exercises. It is scheduled to transition to the forces in 1999 under the Special Operations Command's funding line. The Air Force's Air Combat Command is meanwhile sponsoring a related program, Steel Eagle, for an air- emplaced version.
Destined for installation aboard the USS Invincible, Cobra Gemini will monitor foreign missile launches. (Hanscom AFB photo)
Chemical Detection: The CMO has also run a demonstration mounting a small surface-acoustic-wave system on a Pioneer UAV. According to the source, the system successfully fed chemical-agent detection and quantification data to the UAV ground station in real time.
Radio Frequency: This is naturally the area of most interest to JED's readership. Unfortunately, it is also the area most closely guarded by MASINT program officals. The most the CMO will admit to is involvement in some unique, unintentional RF emissions work, looking at wideband RF signatures. The focus is not traditional electronic intelligence, but to investigate "what other little trails does a radar leave," said the official. But the realm of advanced RF intelligence - be it under the rubric of ESM, SIGINT or MASINT - isn't all that big. The scrupulous researcher can probably glean a good deal about cutting-edge RF MASINT by following the latest pursuits of service and industry labs in specific emitter identification. Essentially, these programs are developing radar- warning-sized tactical receivers that can capture the unique "fingerprint" of any individual threat emitter. This entails, naturally, very precise, very detailed, very processing-intensive analysis.
Specifically, the US Air Force Wright Laboratory's Precision Location and Identification (PLAID) demonstration is applying some very MASINT-like techniques to emitter characterization. The system's narrowband digital-analysis receiver correlates multiple emitter inputs, including the nonintentional -- formally known as unintentional modul ation on pulse (UMOP). Sources of UMOP can include the threat system's oscillator, the power supply or any other ancillary components that yield very fine frequency modulation, and thus a unique signature. (See "US Labs Find the Fine Grain for Fighter ESM" in the September 1997 JED, p. 49.)
For an upstart intelligence discipline only institutionalized within the last five years, MASINT technology is surprisingly mature. The hardware technology is mostly extant. The actual sensors either exist or will exist in some useful form in the near future.
Where the technological hurdles remain are on the back end: disseminating the data and exploiting it in a manner timely enough for tactical utility. One of the problems is the immense volume of intelligence some of these sensors collect. The reams of data collected by a hyperspectral sensor operating across hundreds of bands require a tremendous datalink (even after compression) to reach the ground station. Ground-station software still needs to be developed that can sort, peruse and translate that data into useful information.
The biggest task ahead of MASINT systems developers may be the construction of the necessary signature databases. For a MASINT system to make rapid and accurate identification on the battlefield, these libraries are critical. In the case of the PLAID, for instance, existing threat-emitter databases are useless, since they contain data recorded by older, much lower fidelity receivers. In other cases, such as chemical-agent detectors, whole new paradigms are involved. No one has yet constructed a library of collected molecular structures.
"Right now," said the CMO official, "hyperspectral systems are good at detecting in real time....But we're working on massive signature-collection matrices so that you could near-real-time or real-time - and I mean within seconds - do a type of signature identification." Such an undertaking is at least five years away from any kind of tangibility, he noted.
Whatever technological obstacles face the budding MASINT discipline pale in comparison, however, to the bureaucratic tar pit in which it is mired. The US's mammoth intelligence infrastructure grew up on a set of organizational principles based on the classic intelligence disciplines. Those disciplines are today wellensconced, institutionally. The NSA, for example, wields near-autocratic authority over all US government SIGINT activities. Collectively, the SIGINT activities conducted by the NSA, the CIA, the services and other government entities form the US SIGINT System. The director of the NSA is the designated manager for this system. He backs up this title with a real manifestation of power control of over half the SIGINT budget. He also reports directly to the secretary of defense and serves as principal SIGINT advisor to the secretary, the director of central intelligence and the chairman of the Joint Chiefs of Staff.
In principle, the CMO director is the single manager for MASINT, like the NSA director is the single manager for SIGINT, but in practice the CMO is a weak organization with far too little visibility within the defense establishment. This is through no fault of its own; rather it is because the CMO is tucked away in the DIA org chart. As the 1997 DOD Inspector General's report points out, the CMO director ultimately does not have direct control over the MASINT General Defense Intelligence Program funding: "The result is the Director, CMO, cannot effectively manage MASINT under its current organizational placement. This condition adversely affects the Central MASINT Office's influence in areas such as establishing standards for MASINT systems and interoperability of these systems with systems of other Departments or Agencies."5 The CMO also has no presence on the Military Intelligence Board, but is instead spoken for by the DIA director.
This bureaucratic weakness has in turn compounded MASINT's visibility problem within the US intelligence and military communities. Despite the CMO's programmatic vigor, the Inspector General's investigation discovered that only 43% of the warfighting organizations visited had knowledge of the MASINT requirements submission process and only 13% promoted MASINT use in military exercises.6 The end result, notes the report, is that the CMO has not succeeded in promoting MASINT as an intelligence discipline that provides support to military operations. Part of the problem is that MASINT is ahead of its time - or its bureaucracy. It enfolds so many disciplines that it does not fit neatly into any of the established pigeonholes, and therefore languishes in conceptual limbo.
One novel solution to this dilemma would be to refocus MASINT on its roots -- national intellisence sunport. There are distinct advantages to such an approach. Some critics have argued that the craze for SMO has thrown the technical-collection intelligence agencies, who've always been DOD-leaning in the first place, far too close to the battlefield, to the detriment of strategic intelligence. At the same time, US strategic intelligence requirements are growing, as the proliferation of weapons of mass destruction (WMD) becomes the post-Cold War public enemy number one. US strategic assets are already spreadthin, as the failure to predict India's recent nuclear tests demonstrated.
Both the CMO and the intelligence community at large could be well served by the organization's tackling of the special weapons portfolio. On the one hand, the CMO would counterbalance the growing gravitational pull of SMO on the intelligence agencies. On the other, the CMO would gain higher visibility and a better focus for MASINT collection and production. John Pike of the Federation of American Scientists recommends the combining of the CMO, the Air Force Technical Applications Center, the Forest Green program office (the Air Force MASINT office at the NAIC), the Missile and Space Intelligence Center and the Armed Forces Medical Intelligence Center (biological warfare) into a single, fairly substantial WMD intelligence organization. That could well make MASINT a bureaucratic contender, if not a heavyweight.
"1C21: Intelligence Community in the 21st Century," Staff Study, Permanent Select Committee on Intelligence, US House of Representatives, April 1996, p. 147. 2. "Evaluation Report on Measurement and Signature Intelligence" (Report No. PO 97-031), Inspector General Department of Defense, June 30,1997, p. 2. IC21, p. 153
Cobra Brass, Federation of American Scientists Space Policy Project, http://www.fas.org/spp/military/program/ warning/cobrabrass.htm. Inspector General DOD, p. 10. Inspector General DOD, p. 11