Air University Review, September-October 1977

IFFN

a technological challenge for the '80s

MANY modern weapon systems have a large mismatch between their maximum performance capability and the performance they are actually allowed to achieve. Because current rules of combat engagement normally call for visual identification of a target as hostile before weapon firing can be initiated, many weapon systems do not operate at anything approaching their design capability. A particular and dramatic case in point is a currently operational long-range missile and control system. With multiple target acquisition and tracking and engagement capability at stand-off ranges of up to one hundred miles, this is a potentially potent weapon system. Under the visual rule, however, where it is difficult to see even large aircraft at more than two miles in good weather (identification can probably occur at no more than half this range),^l there is a serious question as to how the user will realistically capitalize on this capability.

A solution to this and related weapon utilization problems is through the development of automated IFFN (Identification, Friend, Foe, Neutral) systems. Indeed, with sensors in satellites, aircraft, ships, and on land and sea floors combining to form a single network of enormous connectivity, the military services of the United States are moving toward a total real-time command, control, and communication capability on a planet-wide basis. The reason for this is that the services have come to realize that it is becoming increasingly difficult, due to weapon proliferation, to answer their most basic question: Where is the enemy?

Korea was probably the last war in which there was anything that might be called a forward edge of the battle area (FEBA), a reasonably well-defined line between opposing forces. With a FEBA, your friends are those on the same side of the line as you, and your foes are those on the other side. Neutrals or noncombatants can occur on either side. If a FEBA exists, IFFN of unknown targets can be made on the basis of geographic location or point of origin. However, the experiences of the United States in Southeast Asia and observations of the recurring Arab-Israeli Middle East conflicts have driven home a reality of modem warfare. It is no longer a set piece, move/countermove, majestic sequence of operations. It is a swirling, lightning-fast, explosive mixture of friends and foes alike, each trying to sort the other out.² The side that does so first will have the advantage, possibly a decisive one.

Even small advantages can be extremely important, as Possony and Pournelle point out with their example of two fighter aircraft, each equipped with "long" range acquisition radars and "long" range air-to-air missiles.³ If "long" means 50 miles to one side but 52 miles to the other, this four percent advantage could mean that one fighter will be detected, acquired, and destroyed before its pilot is aware that he is not alone in the sky. Of course, this is overly dramatic because a 52-mile missile launch would not be performed on the mere basis of a radar track (presumably both sides have the same IFFN problem). An example of this "detection but no identification" problem resulted from the similar appearance on a radar display of the F-4 and the French-British Concorde Supersonic Transport. On test runs between London and Bahrein, Iraq sent up fighter interceptors to visually identify the Concorde because only Israel flies the F-4 in that part of the world.⁴

We can understand, then, the necessity for the visual rule. The only truly positive technique available today for distinguishing between friend and foe (and maybe neutrals, too) is to look at them. To do otherwise is to risk fratricide.⁵ This is not to say there are no alternatives to visual identification. One can use the correlation of the location of an unknown target with the known locations of all friends. The lack of a match might be taken as an indication of foe. Not only is this concept relatively slow, with its implication of the existence of a high-level command and control system that "knows all," but it is not really a positive identification of a foe. There are, however, plans for more responsive identification systems using sophisticated versions of this concept; they go under the generic name of time division multiple access (TDMA). These systems will require large expenditures of money and significant changes in procedural operations, however, and do not easily allow autonomous operation of individual weapon systems.⁶

Another alternative to visual identification relies on explicit procedural methods, e.g., aircraft flying in safe passage corridors specified in time, speed, and altitude. The major objection to this is the obvious inflexibility and the opportunity for the enemy to learn the procedures by observation. (If procedures are changed frequently to defeat such learning, then the problem occurs of ensuring that all friends and neutrals are always informed in a timely way, while still denying this information to the enemy.)

The problem of identifying friends and foes in war is not new. The use of uniforms, flags, and other visual insignias has a long history. Certain special categories of participants in a combat or potential combat area have also developed visual cues to announce their identity or intent, e.g., the red cross for unarmed medical personnel, the white flag for soldiers wishing either to negotiate or surrender, and blue helmets for the troops in a United Nations Emergency Force (UNEF). To ignore or violate the spirit of these insignias has been to invite condemnation by world public opinion or, in the case of using an adversary's uniform, to risk execution as a spy.⁷ Passwords have served as an acoustic Identification, Friend or Foe technique for centuries.

What is new is the need to develop the capability to engage the enemy at long range. It is not desirable to continue to follow Israel Putnam's two-hundred-year-old exhortation at Bunker Hill of "Don't fire until you see the whites of their eyes." Missile weapons, for example, typically have their largest probability of kill (P_k) at near their maximum range, with a plot of P_k vs. launch range to target appearing something like Figure 1. Not to launch such a weapon at long range is not only to accept a degraded P_k but also to increase the vulnerability of the missile weapon platform to counterattack and lose the element of surprise. Long-range identification also provides two valuable threat assessment capabilities: (1) if the identified threat is too potent to handle, the engagement might either be aborted or postponed, if possible, until a friend capable of engagement arrives; (2) if the decision to engage is made, long-range identification may provide sufficient time to set up the optimal attack geometry for the particular friend-foe combination. These comments are not limited to air-to-air combat. The restrictions of the limited visual rules of engagement carry over to air-to-ground (e.g., air strikes against land and sea logistic and combat forces), ground-to-ground (e.g., tank vs. tank battles), and ground-to-air (e.g., surface-to-air missile [SAM] defense sites). The IFFN question drives the engagement decision process in literally all forms of combat.

The visual restriction has, up to now, been imposed because of the disastrous consequences, in our own eyes, of a mistake. To engage and destroy a friend or neutral are viewed as nonjustifiable.⁸ On the other hand, to some the continued adherence to this policy is an unrealistic application of the judicial philosophy that a target is a friend until visually proved to be a foe. Much work in the past has been devoted to the study of the potential for eyeball detection and identification of targets (both with and without artificial aids).⁹ More recently, interest in nonvisual techniques for the noncooperative identification of targets has developed. Some extended comments are in order on what is meant by cooperative and noncooperative IFFN.

A cooperative IFFN system is one that requires targets either to play a responsive role in their identification, upon request from remote observers, or to continuously enhance one or more of their observable characteristics that aid in the identification process. Ground troops that wear uniforms, ships that fly flags, and aircraft with insignia painted on their frames are examples of the latter. These are examples of passive cooperative techniques. The modern radio beacon transponder that broadcasts either clear or coded signals upon interrogation exemplifies an active cooperative technique. The terms "passive" and "active" describe the role of the target observer. A noncooperative IFFN technique, which can be either passive or active, requires no participation by the target in the identification process. Passive noncooperative techniques have the virtue of not emitting and thus of not giving away I the observer's position.

The beginning of active (electronic) cooperative IFFN can be said to have occurred during the Second World War in a parallel (if less dramatic) development with radar. The initial euphoria over the ability of radar to look through darkness, weather, and distance to provide target range, bearing, and speed was soon tempered with the realization that without target identification little could be done but track until visual identification could be performed. This need in the military for a fast, reliable, long-range means for radar target identification triggered the development of the Mark series of IFF radio beacons, culminating in today's Mark XII model,¹⁰ widely used in military aircraft. The civilian counterpart plays an enormously important role in the nationwide Air Traffic Control Radar Beacon System (ATCRBS) radio interrogation network.¹¹ In fact, the civilian ATCRBS and military Mark XII have coalesced into a single entity in the form of the United States Air Force 407L Tactical Air Control System in Western Europe. This system is also known as AIMS, an acronym for ATCRBS IFF Mark XII System, which is itself a sequence of acronyms. Acronyms cubed!

The Mark XII active cooperative IFF is generally considered to be a very reliable system—it is also cryptosecure—but it has some significant drawbacks. Foremost among these is that it is really a misnomer to call Mark XII an "IFFN" system. It certainly has no neutral ¹² identification capability, and its identification of a target as a foe is by elimination, i.e., it is a positive identification system only for friends possessing a working Mark XII. It cannot discriminate a hostile target from those friendly targets which, for a variety of reasons, fail to respond to an interrogation. Mark XII itself does not designate targets that do not answer interrogations as foes but rather as unknowns. Other considerations are required to complete the identification task. Figure 2 shows the nature of this process.

The process shown in Figure 2 is satisfactory in situations similar to the one the United States had in the air war in the northern part of South Vietnam. In the region of the central highlands around the base at Pleiku, north through Da Nang and Hue to the Demilitarized Zone that separated what used to be South and North Vietnam, the American forces enjoyed complete air superiority. This was combined with the presence of forward air controllers (FACs), who provided accurate position and identification information for close-air-support missions against ground targets. However, this total, absolute control of the air would probably not be a reality, for either side, in the potential "conventional" confrontation between NATO and Warsaw Treaty Organization (WTO) forces in Europe.

From the point of view of the United States, several factors combine to make the AIMS Mark XII/visual identification combination an unsatisfactory answer to the total IFFN problem in Europe, in the context of war. The presence of large numbers of the MiG-21 and variable-geometry MiG-23 Flogger will probably deny total air superiority to NATO, even when equipped with the F-16. The recent decision to deploy a wing of F-15 Eagles in West Germany and to add a second wing of F-111s in Britain may alter this evaluation, however. The long-term survivability of FACs in Europe is doubtful in the face of the heavy radar-directed air defenses they would most surely meet. The visibility in Central Europe is generally poor (e.g., less than two miles 20 percent of the time in winter). And finally, there seems little doubt that the Warsaw Pact forces are trained and equipped to maneuver and fight at night.¹³  

Because of these considerations, interest in noncooperative IFFN techniques has grown during the past several years. In 1974, a panel of experts reported to the Army on their survey of the state of the art. More recently, Dr. Malcolm R. Currie, then Director, Defense Research and Engineering, requested that a Defense Science Board (DSB) Task Force be established to study the IFFN problem and make recommendations.¹⁴ In a parallel effort with the DSB Task Force, which was established in early 1974 with members from both government and industry, the Institute for Defense Analyses (a Federal Contract Research Center, primarily funded through the Office of the Secretary of Defense)¹⁵ performed an IFFN technology study for Defense Advanced Research Projects Agency (DARPA). These technologies cover the entire electromagnetic spectrum, from UHF to infrared. Table I lists just a few of the techniques that have been discussed in the unclassified literature. ¹⁶

But problems still exist. The techniques listed in Table I and other noncooperative IFFN technologies are in danger of being compromised by the realities of the extraordinary acceleration over the past two decades of the world arms trade. The possession of a particular type of military hardware is not necessarily an indication of nationality. This proliferation of weaponry, due primarily to the willingness of the United States and Soviet Union ^l7 to sell even their most advanced developments (short of nuclear weapons) to Third World countries, has reduced most noncooperative techniques to the level of target classifiers,* as opposed to friend-foe identifiers. In the Pakistani-Indian War of 1965, U.S. weapons appeared on both sides. The 1974 invasion by Turkey of the island of Cyprus, under Greek control, is a more recent example of the IFFN problem caused by the widening distribution of weapons. Greece and Turkey, both members of NATO, used U.S.-made weapons against each other, and weapon type implied nothing about the nationality of the possessor. A similar situation would face the United States if it should engage in a war with the Organization of Petroleum Exporting Countries (OPEC) oil cartel, since the Middle East, particularly Iran, has purchased enormous quantities of American weapons.¹⁸ Some appreciation of just how large this weapon proliferation problem has become can be gained from Table II, which shows the distribution, in fixed 1970 U.S. dollars, of total world military expenditures in 1964 and 1974.¹⁹ While the U.S., England, and France all experienced almost insignificant increases over this ten-year period, the Third World actually gained on the United States and Soviet Union in total dollars spent and surpassed all others in growth by more than doubling.

*The definitions used are as follows: Classification--determination of equipment model and type: General Identification--determination of the nationality of the armed force operating the equipment; Specific Identification--determination of the particular military unit, item or vehicle (by tail number, serial number, etc.) being observed.

The increasing spread of weapons around the world certainly bodes potential ill for humanity in general, and it seems almost inappropriate, by comparison, to observe the difficulties this spread causes for IFFN. The late Walt Kelly's famous line from the comic strip Pogo comes to mind when trying to express the proliferation problem: "We have met the Enemy and He is Us." Certainly Kelly wrote this in a different context, but the statement has new relevancy in view of the far-flung distribution of weapons.

Each separate technology for noncooperative IFFN takes advantage of the individual and special characteristics of the target signal (called the "signature") available to a remote observer. These special nuances constitute, in many cases, classified information. However, the nature of many of these signatures and the general limitations inherent in them that reduce their usefulness, because of the spread of weapons or because of the complex signal processing they require, can be found in HR³, a signature listed in Table I and discussed in the open literature. HR³, high range resolution radar, is an active noncooperative technique because the target must be illuminated by an observer's wideband pulse radar. Large bandwidth (on the order of hundreds of megahertz) is required to achieve a range resolution on the order of feet. This allows the radar receiver to distinguish the echoes of the individual locally dominant scattering sites on a complex target.

Viewed on an A-scope display (echo signal amplitude vs. time or, equivalently, range), the HR³ signature is a "range profile" signature consisting of a sequence of peaks, corresponding to the significant scattering sites on the target. Figure 3 shows a typical display of the HR³ signature. An important characteristic of this signature is that azimuth information is lost,²⁰ with the signature peaks appearing in positions corresponding to the projections of the dominant scatterers onto the radar line of sight (RLOS). Since the information in an HR³ signature is in the relative strengths and positions of the peaks (these "geometrical" features are different for different target types), there is clearly a viewing aspect dependency inherent in this signature. Such a signature is able to provide only general identification, even when the national ownership of the target type is limited. If there is a wide distribution of the target type around the world, then the capability of the signature is reduced to providing only classification.

Because of this reduction in the capability of a signature like HR³ to perform IFFN, there is interest in searching for "fine structure" in signatures. Success in this search might lead to the ability literally to "fingerprint" each copy of a weapon system at the time of manufacture. Fingerprinting can be thought of as having two distinct origins: fine structure in the noncooperative signature due to either (1) intrinsic variations among copies of the same weapon system or (2) intentionally introduced variations, i.e., a built-in "serial number" in the signature.

The concept of "signature fingerprinting" introduces considerations of military intelligence in a direct and immediate way. To use a catalog or library of fingerprinted signatures effectively, the geographical deployment of particular copies of a weapon system becomes essential information. This information must be kept timely to be useful and can be degraded by such occurrences as secondary arms sales by the original purchaser, attrition from accidents, wartime loss and wear-out retirement, and redeployment to new locations. This kind of intelligence information may be very difficult to obtain.

Another serious problem with fingerprinting is that it allows the weapon system owner to possess the signature fingerprinting mechanism, even if unknowingly. If the fingerprinting mechanism details are at all compromised (i.e., stolen or "leaked"), a fingerprint could possibly be obliterated, reduced in visibility, or, worst of all, altered to appear as a friendly fingerprint. Fingerprinted signatures may prove to be highly perishable and a double-edged sword.

The problems in creating a useful noncooperative target signature library do not really disappear even in the case of nonfingerprinted weapons. For other than U.S. made weapons, noncooperative target signatures obviously have to be obtained by the method of looking, on a "target of opportunity" basis, at a representative member of the target class of interest. This can be risky for active noncooperative target signatures. For example, radar illumination of high performance potential adversary fighters to obtain their HR³ signatures conceivably are provocative acts (depending on when and where) and have clear potential for political exploitation. ²¹

Noncooperative target IFFN technologies have an existence of their own, quite apart from any particular weapon system. However, in responding to the question, "Will the use of Technology X result in a significant increase in the effectiveness of System Y?" the IFFN technology and the system cannot be decoupled. This unfortunate reality greatly complicates the already difficult task of evaluating the capabilities of just the technology without worrying about how to interface it with a system. It is not at all difficult to construct a fairly long list of important considerations in an IFFN technology assessment, including the sensor and associated signal processor. (Table III shows some of these issue areas.) Adding the additional items of "Weapon System Interface Problems" and "Determination of Enhanced System Effectiveness" requires that the specifics of the weapon system and its operational environment be considered, too.

More will be said about the second of these two items, and the interface problem is dismissed with a brief platitude that admittedly offers no immediate help: "When the horse has been stolen, the fool shuts the stable." By analogy, in the interfacing of IFFN technology with an operational weapon system (in all probability designed with IFFN as a low-priority consideration, if considered at all), the resulting required "fixes" may actually change the character of the system. Interface control and data paths to and from a candidate IFFN technology and the system may not be readily accessible or even exist. Extensive and costly modifications may be required to build the interface. For existing weapon systems, this situation can only be lived with; but for new systems still in the conceptual stage, the time to think about IFFN is right now, i.e., lock the barn door while the horse is still in the stable munching hay. In its most extreme form, this viewpoint is best expressed by those who would require the preparation of an "IFFN Impact Statement" by the advocates and designers of any proposed new weapon system. With this requirement, no new weapon system would be allowed to proceed beyond the basic research and development stages until it had been demonstrated how the IFFN problem would be addressed.

Table III. Some important IFFN technology issues

• signature description

• viewing aspect dependency

• cost (initial, maintenance)

• size/weight/power requirements (s)

• sensor parameters (frequency, bandwidth, range and angular resolutions, field of view)

• error rate(s) and level (s) of confidence for signal processor output(s)

• vulnerability of sensor and/or signal processor to countermeasures

• sensor input (antenna) requirement

• simultaneous multiple signature capability

• feasibility of obtaining target signatures

• signal processing requirements (memory size, timing, data rates)

A natural result of such an impact study would be answers to the question of what the payoff due to the inclusion of IFFN is (i.e., the determination of enhanced system effectiveness) as compared to the system effectiveness without the proposed IFFN technology. To carry out this kind of analysis requires some measure of system effectiveness that quantitatively evaluates system performance. Unfortunately, a single effectiveness measure does not exist that applies to all systems; also these many measures are all functions of such widely variable considerations as system cost and mission requirements.

For example, if we consider two systems, one "tactical, low cost" and the other "strategic, high cost," distinctly different measures of effectiveness are appropriate. For the first kind of system, an example of which might be a light tank, an economic exchange ratio is a reasonable measure. Tanks can be made on a mass production basis, and even if their survivability in war is not particularly good, that may be permissible if during their lifetime they cost the enemy more than our cost to replace them. The larger the ratio of adversary cost ²² to our replacement costs, the more effective is the tank weapon system.

Such a measure of effectiveness is surely not an appropriate one for a nuclear aircraft carrier, an example of the second kind of system. Nuclear carriers are enormously expensive, very low production rate weapon systems, and, if anything, a cost exchange ratio works to the advantage of an adversary. (Several missiles delivered by KOMAR or OSA class boats are far cheaper than a carrier with its complement of aircraft.)²³ A meaningful measure of effectiveness for this kind of system would include not only the capability of the system to damage the enemy but also the probability the system survives a complete mission with the ability to undertake a new one.

The value of a noncooperative IFFN aid to a weapon system is directly related to how much it improves the system effectiveness measure. The manner in which such aids will influence these measures is by increasing the range (beyond the visual) at which target identification can be achieved. The first step in performing an IFFN enhancement analysis, then, is that of answering the question, "How much sooner can the system identify a target with an IFFN capability than without it?" Even this first-step analysis, in its most elementary form, must consider various complications introduced by the interaction of the mission situation of the system and the particular nature of the target signature. For example, in his scholarly analysis W. D. White states that if one compares aircraft combat loss rates (usually given as the number of aircraft lost per 1000 sorties) over the long historical period from World War II to the Yom Kippur War" no evidence exists that suggests a decline in the survivability of tactical warplanes over a modern, conventional battlefield.²⁴ A new U.S. Army weapon that might change this evaluation is Stinger. Stinger is a small (21 pound) shoulder-launched SAM. Using passive infrared guidance with proportional navigation, it is intended to give the mobile ground soldier the capability to engage low flying, high speed (up to Mach 2) targets.²⁵ Stinger is equipped with an IFF aircraft interrogator.

In discussing how a noncooperative IFFN technology aid might be used with a Stingerlike weapon, one must keep in mind that there has to be a balance between the aid and weapon in such aspects as mobility, size, and cost. For a long-range area defense weapon like the Hawk missile, it makes sense to think of using an aid that incorporates a sophisticated acquisition radar. For Stinger, something else less ambitious is more reasonable; for example, a low cost, binocular-size visual aid. ²⁶ More specifically, a reasonable scenario for a small, portable Stinger-like SAM integrated with a noncooperative IFFN visual aid might correspond to Figure 4.

Figure 4. Battlefield geometry for Stinger

The battlefield SAM is located some distance from a hill, behind which it is known or suspected that potential hostile aircraft will appear. The SAM soldier searches the airspace above and beyond the hill with his IFFN visual aid. The aircraft is assumed to be in level, constant-speed flight, radically inbound toward the SAM. With this battlefield geometry or, in fact, with alternative geometries, one could write formal equations (but not here!) relating the variables of Figure 4, defined as follows:

Finally, by making some plausible assumptions of the unaided visual IFFN capability of the SAM soldier and knowing the maximum LOS detection range of the candidate technology, one can calculate a quantitative statement of the improvement a particular technology brings to a weapon system. If the Army has not done this for Stinger, it should.

Some think that the best way to advance the state of the art is by continually seeking new sensor phenomena that avoid most of the faults of their predecessors while introducing no new major difficulties. An alternative path takes the point of view that there presently exist a substantial number of distinct sensors, together covering an enormous spectral width but with little knowledge to guide their effective cooperative interaction. This second path is the theme here, the message being that some of the dollars currently being spent on searches for new phenomena might be better spent on the effective integration of sensors already available.

However, before examining the integration question, one should consider the computational aspects of the signal processing load implied by the IFFN technology. This is because the sensor signal processor is the next level of sophistication beyond the sensor, with a multisensor architecture coming after that, as shown in Figure 5. Whatever the physical nature of a sensor, the information provided by it (the "signature") is useful only after at least some minimal processing. A significant (possibly a major) fraction of the cost of an IFFN system will not be represented by the sensor but by the electronic signal processing that will back up the sensor.²⁷ A critical factor in the credibility of present and future IFFN systems will be the signal processing package (e.g., reliability, size, maintainability, power requirements, and speed). Recent electronic device advances need to be explored in depth for their potential impact on the total IFFN system package. The level of risk and feasibility for "shrinkage" of cost, speed, size, input power, etc., via new device technology needs assessing for the various current IFFN sensors. There is little question to the belief that we are, today, not even close to the fundamental quantum limitations on the speed and size of electronic signal processing devices. ²⁸

The next level of sophistication beyond the signal processing logic and hardware that interface directly with a sensor is that of multisensor integration. This is a big step and, until recently, one with a dismal record. Multisensor integration is an immediately convincing systems approach to getting more performance out of a collection of sensors than any single one of them can provide. This is the so-called "synergistic" effect achieved when the multiple sensors feed into some kind of high-level "parallel" processor. And that is the troublesome part of this systems approach, in that parallel systems are so poorly understood. The following quotation from the introduction of Minsky and Papert's elegant book is appropriate:

Neither the outsider nor the computation specialist seems to recognize how primitive and how empirical is our present state of understanding of such matters. We do not know how much the speed of computations can be increased, in general, by using "parallel" as opposed to "serial" or "analog" as opposed to "digital" machines. We have no theory of the situations in which "associative" memories will justify their higher cost as compared to "addressed" memories. There is a great deal of folklore about this sort of contrast, but much of this folklore is mere superstition; in the cases we have studied carefully, the common beliefs turn out to be not merely "unproved"; they are often drastically wrong.

The immaturity shown by our inability to answer questions of this kind is exhibited even in the language used to formulate the questions. Word pairs such as "parallel" vs. "analog" are used as if they referred to well-defined technical concepts. Even when this is true, the technical meaning varies from user to user and context to context. But usually they are treated so loosely that the species of computing machine defined by them belongs to mythology rather than science.

Now we do not mean to suggest that these are mere pseudo problems that arise from sloppy use of language. This is not a book of "therapeutic semantics!" For there is much content in these intuitive ideas and distinctions. The problem is how to capture it in a clear, sharp theory.²⁹

These are not discouraging words but rather signposts that should be recognized as pointing the way to a potentially fruitful area of IFFN research and development. Indeed, there are many military environments, today, rich in distributed, multisensor systems.

The previously mentioned AWACS in Europe, designed to coordinate NATO forces in a confrontation with Warsaw Pact forces, comes immediately to mind. In addition, the Army has long been experimenting with battlefield sensor systems, with the "McNamara Wall" (the U.S. experiment in Vietnam with an electronic Maginot Line) probably the best known example. More recently, the Army is now developing the Remotely Monitored Battlefield Sensor System (REMBASS), made up of widely scattered Unattended Ground Sensors (UGS). ³⁰ These UGS will come from a mix of sensors, including magnetic, seismic/acoustic and infrared detectors. Somewhat along the same lines is the Army interest in a field-artillery acoustic location system. While REMBASS is for use against ground personnel and vehicles, the artillery system is intended to support the rapid development of accurate counterfire against hidden mortar and gun emplacements. This is a technique used in World War ³¹ but with new sensors and computer processing, the Army thinks it can defeat the problems of old, such as echoes off nearby hills.

An even more fantastic example of a military multisensor system was dramatically thrust into public view in 1975, with the disclosure of the CIA adventure called "Project Jennifer." ³² In 1968, a Soviet missile submarine suffered an explosion while recharging its batteries on the surface of the Pacific and sank to a depth of 16,000 feet. The noises of its break up were detected by the UGS scattered on the ocean floor by the U.S. Navy. By using time of arrival (TOA) processing techniques, American authorities knew the location of the doomed boat to within ten square miles, while Russian search ships had no real idea of where to look. Once the Russians gave up, the now famous Howard Hughes ship, the Glomar Explorer, pulled off what certainly must be admitted to be a technological pièce de résistance, no matter what one may think otherwise about the affair.³³  

Any credible proposal for a multisensor noncooperative target IFFN scheme must answer at least two questions: How will the multisensor system combine the individual sensor inputs to arrive at a final target type decision? What is the quantitative pay-off for the additional complexity of a multisensor architecture, as compared to a single sensor? That is, Is more better? These are absolutely essential questions, and any multisensor proposal that avoids them just has not been thought about long enough, hard enough. A third question should also be added: How will the multisensor system handle the correlation problem when multiple targets are observed simultaneously? The problem of correlating a sensor measurement with a target does not exist when just one target is observed, but with multiple targets and passive sensors, ³⁴ it is not clear how the sorting of targets and measurements can be done.

Finally, there is the issue of the psychological interaction between a dispassionate target-classifying machine and a combat soldier in a stressful environment. Should the automatic pattern-recognition signal processing logic always make a positive decision ("friend," "foe," or "neutral"), or should the fourth possibility of "unknown" also be included as an output? And in any case, once a decision is made, should the probability that it is correct (i.e., the "confidence level") also be an output? Does the presentation of uncertainty to a human observer in a situation that is often life or death add to or detract from the overall effectiveness of the IFFN system?

THE IFFN issue is a "sleeper" technological challenge for the 1980s. To be second in this area would be a technological surprise on the United States with enormous repercussions. As Dr. George Heilmeier, Director of the Defense Advanced Research Projects Agency, wrote in "Guarding against Technological Surprise,"³⁵ to be so surprised is not a matter of coming in second. It is to lose.

University of New Hampshire

Notes

1. The author attended a demonstration of the Air Force precision flying group, the Thunderbirds, at Pease AFB, Portsmouth, New Hampshire. Even when flying in a group of six aircraft, in beautiful weather, the team's T-38 Talons were extremely hard to track when more than a mile distant, even though their approximate light path was known beforehand.

2. See Aviation Week & Space Technology, June 30, 1975, p. 12, for example; an Egyptian Air Force gun camera picture taken by a MiG-21 Fishbed (in the October 1973 Yom Kippur War) on the tail of an Israeli Mirage, which in turn is on the tail of another MiG-21. Also, the testimony of Major Steve Ritchie in 1974 before the Tactical Air Power Subcommittee of the Senate Armed Services Committee about combat between the F-4 and the MiG-21 is interesting. As reprinted in James W. Canan, The Superwarriors: The Fantastic World of Pentagon Superweapons (New York: Weybright and Talley, 1975), "The MiG-21, compared with the F-4, is about half the size, it leaves very little smoke, it is hard to see. . . "Finally, in Thud Ridge, Colonel Jack Broughton, USAF (Ret), describes the concern of F-4C pilots when, in a shootout against MiG-17s over the Red River in North Vietnam, accompanying F-105 Thunderchiefs began launching Sidewinder, infrared seeking missiles. From some angles, the F-4C and the MiG-17 are visually similar, and an accident was a real possibility.

3. Stefan T. Possony and J. E. Pournelle, The Strategy of Technology: Winning the Decisive War (New York: Dunellen, 1970), pp. 38-39.

4. See Newsweek, August 18, 1975, p. 13.

5. For example, as reported in Newsweek, November 26,1973, p. 26, Soviet-built SAM-6 missiles were effective against Israeli aircraft in the Yom Kippur War. Yet, because of poor coordination between Egyptian aircraft and air defenses (i.e., no friend-foe discrimination), 40 of the 120 aircraft lost by Egypt were shot down by the Egyptians themselves.

6. Such a correlation system will be part of the Airborne Warning and Control System (AWACS) for NATO in Europe. The FY76 funding request included $199 million to continue AWACS development, including work on the positive identification correlation system (Aviation Week & Space Technology, March 17, 1975, p. 28).

7. These constraints are quite strong. For example, even as they initiated an undeclared war at Pearl Harbor, the Japanese air forces did not fake insignia.

8. The only possible exception to this would occur if a friend mistakenly attacked one of his own. Then the right to self-defense takes priority, and the one attacked, by military doctrine, has the right to take any action, including destroying the attacker, to survive.

9. H. H. Bailey, "Target Detection through Visual Recognition: A Quantitative Model" (Santa Monica: Rand Memorandum RM-6158-PR, February 1970).

10. The Mark series evolved sequentially up through the Mark V at the end of World War II. Just before the Korean conflict, the Mark V was modified, but since there had already been an experimental Mark VI in 1945, just before the end of the war, it was not clear what the designation should be. Thus, the mod Mark V was called the Mark x, where x denoted the unknown. Soon, however, it became the Mark X, and subsequent designations started from there. There never was a Mark VII, VIII, or IX!

11. R. C. Renick, "An Improved ATC Radar Beacon System," Proc. IEEE, March 1970, pp. 413-22. Special Issue on air traffic control.

12. The concept of neutrals in a combat area is one that, at first glance, appears to be ludicrous. But given the highly political nature of recent wars (Korea, Vietnam, the Middle East), the idea of large numbers of neutrals is credible. For example, there was a great deal of civilian air traffic in South Vietnam all during the presence of U.S. air forces there; also the UNEF troops at Sharm el-Sheikh in the Sinai Peninsula, positioned there after the 1956 Suez War until ordered out just before the June 1967 war; most recently, the American contingent of civilian technicians sent to monitor the integrity of the 1975 Sinai accord between Egypt and Israel.

13. The Yom Kippur War can be used to support this statement. Both the Egyptian and Syrian armored forces (Soviet-equipped) possessed active and passive infrared night vision devices, and they used them to great effect against Israel (the Syrian tank drive on the Golan Heights and the Egyptian armor night crossing of the Suez Canal).

14. The Department of Defense, Program of Research, Development, Test and Evaluation, FY1976, Statement by Dr. M. R. Currie, DDR&E, before the House Armed Services Committee, February 21, 1975.

15. H. Orlans, The Nonprofit Research Institute (New York: McGraw-Hill, 1972).

16. The source of information for Table I is Aviation Week & Space Technology, January 27, 1975, p. 121, except for HR ³ (high range resolution radar), which is from D. Howard, "High Range Resolution Monopulse Tracking Radar," IEEE Trans-Aerospace and Electronic Systems, September 1975, pp. 749-55: J-TIDS/TDMA (time division multiple access) which is from C. E. Ellingson, "Performing IFF with ICNI," Mitre Report MTR-1773, July 1970; Vectors (Hughes Aircraft Company), Winter 1974-75, pp. 18-21; and Aviation Week & Space Technology, January 20,1975, p. 51: wide-band doppler in "Monostatic Tracking Radar Imaging Theory for Rotating Point Target Models with Various Bandwidth and Coherence Conditions," SURC Report TN75-139, June 1975; and harmonic radar which is from R. O. Harger, "Harmonic Detection and Imaging Radar Systems for Nonlinear, Near-Ground, In-Foliage Scatterers," IEEE Tans-Aerospace and Electronic Systems, March 1976, pp. 230-45.

17. England and France are also active arms dealers. See Time, March 3,1975, pp. 34-44. A major factor in the spiraling increase in the arms trade is the rapid rise in the costs of weapon research, development, and production. To enable an economic number of copies to be produced, the weapon market has been expanded from internal consumption to sales abroad. Indeed, it is not uncommon for new weapons to show up in the arms inventories of the buying countries before they do in those of the selling countries! The U.S. antitank Tube-launched, Optically-tracked, Wire-guided (TOW) missile is a recent example.

18. For example, Iran has large numbers of the F-4 Phantom and F-14 Tomcat, and Saudi Arabia has the F-5 Freedom Fighter and, as does Kuwait, the Hawk Air Defense Missile System.

19. See Disarmament or Destruction? Armaments and Disarmament, Stockholm International Peace Research Institute, May 1975, the source for Table II.

20. HR³ signatures obtained with monopulse radars can do better by placing the scatterers on one side or the other of the radar line of sight (RLOS) by using the polarity and amplitude of the available angle video signal. See D. Howard, "High Range Resolution Monopulse Tracking Radar," and D. R. Rhodes, Introduction to Monopulse (New York: McGrawHill, 1959). The result is a crude, distorted two-dimensional "image" of the target.

21. See Khrushchev Remembers, commentary by E. Crankshaw and translation by S. Talbott, Little-Brown, 1970 (volume 1) and 1974 (volume 2) for Nikita Khrushchev's description of Soviet psychology after the 1960 U-2 incident, in what might be a modern classic example of an information gathering mission "gone wrong" and the resulting intense political repercussions.

22. The total adversary cost is the sum of two costs: the damage cost caused by the tank and the cost actually to destroy the tank.

23. The KOMAR is a Soviet 75-ton, 40-knot coastal defense PT boat with two Styx missile launchers. The Styx is subsonic, surface-to-surface with a boat-launched range of about 13 miles; it carries a 1000-pound high-explosive warhead. The OSA is a 160-ton boat with four Styx launchers. See R. D. Colvin, "Aftermath of the Elath," United States Naval Institute Proceedings, October 1969, pp. 60-67, for a vivid description of the cost exchange ratio experienced by the Israeli Navy when KOMAR boats in Port Said sank the 1700-ton destroyer Elath in 1967.

24. W. D. White, U.S. Tactical Air Power: Missions, Forces, and Costs, The Brookings Institution, 1974. White estimates the overall Israeli loss rate in October 1973 as about 8, as compared with 6.5 for the Allied Tactical Air Force during the last seven months of World War II Europe.

25. Aviation Week & Space Technology, March 17, 1975, p. 83. Photographs of live warhead tests of Stinger against a helicopter on the ground and an in-flight aircraft can be found in Aviation Week & Space Technology for December 1, 1975, p. 15, and September 15, 1975, p.19, respectively.

26. If the reader is willing to indulge for a moment in some speculative science fiction, such an aid might be visualized as a hand-held, electro-optical device that is pointed at the suspected target. At the push side-mounted button, the "subharmonic monotone phase" target signature is captured and processed by a computer microprocessor. If the signature is that of a hostile, a red X is projected over the viewing field—challenge to the technologists!

27. An obvious exception to this would be IFFN systems incorporating a radar sensor (as opposed to IR sensors and other intrinsic emission detectors, for example). And even in the radar case, the enormous signal processing load in a multielement phased array radar system might require a financial investment exceeding that of the radar itself.

28. R. W. Keyes, "Physical Limits in Digital Electronics," Proc. IEEE, May 1975, pp. 740-67.

29. M. Minsky and S. Papert, Perceptrons: An Introduction to Computational Geometry (Cambridge: Massachusetts Institute of Technology Press, 1969). 

30. Electronics, May 13, 1976, pp. 29-30.

31. Vectors (Hughes Aircraft Company), Winter 1975/76, pp. 6-10.

32. Time, March 31, 1975, pp. 20-27 and December 6, 1976, p. 23.

33. One legal objection to Project Jennifer is that the U.S. may have violated the law of the sea in clandestinely salvaging a vessel of another country in international waters.

34. A radar sensor inherently solves the correlation problem, of course, but an infrared sensor, alone, for example, has no way of determining how many separate targets are in its field of view.

35. George H. Heilmeier, "Guarding against Technological Surprise," Air University Review, September-October 1976, pp. 2-7.

The United States faces a state of insecurity in the future unparalled in our history. This grows out of the simultaneous existence of two basic factors. First, the juxtaposition in the world of the future of our great slow moving democracy, dependent for decision on the slow crystalization of majority opinion, side by side with powerful secret dictatorships, potentially capable of rapid aggressive action with little or no warning. The second basic factor is the tremendous recent advances of technology, making quite possible the delivery, relatively instantaneously, over vast distances, of powerful blows of such destructiveness as to be potentially decisive. Times have changed—and very much for the worse for us!

Major General Muir S. Fairchild
Editorial, Air University Quarterly Review
Winter 1947

Paul J. Nahin (M.S., California Institute of Technology; Ph.D., University of California at Irvine) is Assistant Professor of Electrical Engineering, University of New Hampshire, Durham. His ten years of industrial experience include the digital logic design of the Gemini-Apollo space telemetry simulator at the Eastern Test Range and analyses of various IFF coding schemes for 407L and AWACS. From 1973 to 1975 he was on the staffs of the Institute for Defense Analyses and the Center for Naval Analyses, Arlington, Virginia and a postdoctoral fellow at the Naval Research Laboratory, Washington, D.C. His current research activity is computer processing of television images.

Air & Space Power Home Page | Feedback? Email the Editor