THE MESH AND THE NET

Speculations on Armed Conflict

In a Time of Free Silicon

MARTIN LIBICKI

Chapter 3


TOWARD AN INFORMATION CORPS

Technology, used correctly, begets doctrine; doctrine begets organization. To the extent that tomorrow's military power is defined by expertise at information rather than the application of force, military superiority may flow to those organized for the former task rather than the latter one.

Today's relationship between weaponry and information resembles the relationships among weapons systems and other supporting elements such as command and control, logistics, and personnel. Operations sit atop; all else supports them. Current weapons have accommodated the information revolution by taking advantage of additional data inputs, but the military remains organized around units of force.

This architecture may soon become obsolete. Tomorrow's winners may build their forces around a central information processing core. Such a core would launch information probes into the media of war (that is, into ground, air, sea, or space arenas, or the spectrum per se). Multiple sensors under various levels of control would gather, transform, fuse, and harness the returning stream, convert it into threat identification, then fire control solutions, and then ladle results in strategic synchrony directly to fire-control units or indirectly to operators.

The traditional relationship between information and force would be turned on its head. Information would no longer serve units of force. Platforms would become vehicles for transporting sensors and missiles. As fewer sensors ride on platforms and more missiles are delivered from unmanned locations (see the section below) the predominance of global information loops will increase. Thus the relationship between information and weaponry is reversed. Rather than information being a service to the weapon, the weapon is the dispatch mechanism slaved to the Mesh. Units of force would be fire support for information systems. Changes in organization imply changes in relationships and status. Current military structures are built around legions of operators served by lesser communities, such as information, logistics, engineering, and communications. Although "lesser" is not meant pejoratively, in any unit which combines such disciplines, operators take command. Moreover, although officer career tracks are similar up to a certain level, operators clearly make up a much higher percentage of the Services' top ranks (major generals, rear admirals, and above) than they do of their overall officers corps. In the Air Force, for example, a quarter of all officers, but over ninety percent of flag officers are fliers. Were information warriors assigned to their own organizations (be they corps, services, or commands), their relationship to the whole would undergo a concomitant and perhaps necessary adjustment.

Rationalizing a Corps

The basic argument for a separate Information Corps and an associated command structure linking operations and intelligence is that it would facilitate effective joint operations, promote the information revolution in warfare, unify the disparate information elements and give them an identity, create a common ethos for information warriors, and provide a unified interface with civilian information infrastructures. It would also provide greater appreciation for the role of information warfare.

Jointness: The farther platforms can see and shoot, the larger their battlespace, and the more service-specific battlespaces intersect with each other. Aircraft of the Navy and Air Force now use the same Air Tasking Order. Data collected by Air Force assets guide Army movements. National sensors alert anti-tactical ballistic missile forces of enemy launches. All the services use the same satellite systems. If nothing else, the sheer rate of growth in the volume and variety of data collected makes the construction of interoperable, or single, information systems all the more imperative.

The information jointness problem bespeaks an important transition in how wars are fought and the diminished local ties between seeking and shooting. Today the two usually are closely linked. Although prepped by intelligence reports, a tank must both find and kill the target itself. Yet other forms of warfare have already experienced the separation: strike operations are planned from externally collected data; anti- submarine warfare operations use an elaborate localizing program prior to administering a coup de grace. JSTARS and AWACS support an efficient cue-and-pinpointing system. The advent of precision-strike systems that use both absolute and relative positioning (that is, latitude, longitude, bearing, range, course, and speed) is at hand. The growing proliferation of sensor systems implies that the targeting systems of tomorrow must be able to fuse data collected from a wide variety of sources. Such fusion means that seamless interoperability is being demanded for missions ranging from single-shot targeting all the way to situational awareness by CINCs.

To illustrate the value of an integrated perspective consider how a hypothetical Unmanned Aerial Vehicle (UAV) sensor package and might be developed -- not only its hardware, but also its software, communications, integration with other data units, and most importantly its doctrine and concept of operations.

The use of UAVs, as all services recognize, can help warfighting. So thinking, each service could develop a package to fit its own mission profiles and support its own platforms. Yet data coming down from UAVs would more logically go to common data receptors and there meld with other joint data collection assets including ground-based sensors, higher-altitude aircraft, and space sensors. To the extent that each sensor package performs its own on-board processing, it may wish to take advantage of common neural training regimens and pattern recognition tools. Data from the various sensor packages -- which could come from any of the services -- have to be analyzed in real time to determine where follow-on data collection efforts have to be focused, or whether and when fire control solutions have to be generated. The interoperability requirements of such a package are therefore demanding.

The need for interoperable information systems has been widely recognized by the senior leadership within DOD. In 1993, former Secretary of Defense Les Aspin observed in a graduation address at the National Defense University, "Most of our systems for the dissemination of intelligence imagery cannot talk to each other." The principal joint command and control initiative (C4I for the Warrior) is almost exclusively about interoperability. It mandates that all new information systems must be able to communicate jointly. Unfortunately, history suggests that after-the-fact standardization frequently leads to unsatisfactory results. Why?

Returning to the UAV sensor package, development by different platforms groups increases the possibility that each system stands alone, making complete data fusion that much harder to achieve.

An Information Corps is an alternate route to data integration. Instead of having the services and DOD agencies (and the multiple communities within them) attempt to merge information collection and dissemination systems, the functions would be carried out by a single organization that operates under a unified doctrine and a single command. Data would be standardized from the start; internecine politics that allow components to agree to disagree would be, if not eliminated, then substantially muted. What would otherwise be a conflict between the need for innovation in data collection, and the subsequent need to report only that which has been standardized, would be muted as well. Successful doctrinal innovations would be integrated into the whole much earlier in their development.

A related rationale emerges from the emphasis on Joint Task Forces (JTFs) that are expected to characterize an increasing share of tomorrow's fighting packages. Such organizations usually are made up of a chunk of this and a chunk of that. To work smoothly most chunk commanders (and key staff members) ought to know each other beforehand. A coterie of information warriors whose specialty is preparing the battlefield image but who are attached to different operating units is already integrated. Acting as the glue, they can integrate far more fine- grained units in precisely that area where interoperation is most important: information.

Innovation: Predicting the demise of the platform is far easier than having operators accommodate this demise. Left to themselves people tend to apply technology in ways which conform to their basic world-view--warriors are no exception. Thus innovations in equipment or doctrine which threaten such an order are likely to be resisted by operators. Granted, no one questions the overwhelming relative superiority of the U.S. Armed Forces, and for that reason our manned platforms would logically be the last to be threatened. However, potential competitors would be foolish to challenge our dominance by a strategy that copied our force structure. Forces built around information systems constructed from commercially available components, however, would pose a more serious threat -- one which contests our reigning paradigm. Thus, it would be far more attractive to challenge us.

Although an Information Corps may not be inherently more innovative than the Services, it is more likely to pursue the kinds of innovations that accord with the logic of the information revolution. Left to themselves, the Armed Forces will incorporate information into weaponry, but with information technologies as platform support rather than with platforms as fire support to an information mesh. An Information Corps would take an entirely different approach from the outset, emphasizing the information mesh as central. Constituent elements and doctrine for such a mesh would be evaluated on their ability to locate, track, and evaluate objects and events so that they may be passed for conversion into fire- control solutions and servicing. Such a service or corps would be an institutional advocate for a paradigm shift, and would, by its advocacy, better prepare for a threat which comes from a different direction.

Unity: The common argument against creating a completely new organization is that its planned functions are all being done by someone else. When this issue is raised, however, the composition of the group varies widely: the Director for Command, Control, Communication and Computer Systems (J-6) on the Joint Staff, the Defense Information Systems Agency, the Defense Mapping Agency, the Space Command, and intelligence agencies -- all without going into the services. Under the last are specific functions such as command and control, electronic warfare, meteorology, oceanography, automated data processing, and high-information platforms such as Aegis, AWACS, JSTARs, and UAV contingents. Other functions which technology may soon enable are not even listed for obvious reasons; when they do emerge, the soup will be even thicker. This is just the point. The various sub-communities in the information-based warfare community see themselves as disparate players. Each relates to one or two others at most, and they all lack the common unifying doctrine of operations. Information warriors are more than simply communicators, data processors, or intelligence agents. They are all part of a global structure that would become apparent with the creation of an Information Corps.

Culture: A related reason for integrating various DOD informational elements into a single corps is to provide information warriors with status, culture, and an ethic. The issue of respect is relatively straightforward. As information becomes more important, so is cultivating the ability to develop and manipulate it. DOD needs to attract these people not only as contractors but more importantly as operators. Successful military organizations must deploy not only superior information systems, they must also be able to fix, adapt, and maintain them in battle in real time. Yet an aspiring officer today would be advised to specialize not in information but in operations. Even the Air Force -- the most information-intensive service -- is oriented toward its fighter pilots just as the Navy is to ship and submarine drivers and naval aviators. Top echelons in other specialties such as administration, material management, and command and control are often assigned from the ranks of operators. This procedure makes sense if various specialties call for similar skills and the best are attracted to operations; an elite is an elite regardless of what it does, and it could as easily be mergers and acquisitions. However, if the skills required to be a good information warrior are different from the qualities and ethos needed to be an operator or these skills require long, specialized training, then such logic makes less sense. The best people avoid information; those who remain do not get the consideration their views deserve.

An Information Corps offers the possibility of separate and more appropriate training and career management as well as an ethos for an information warrior. As computers get more sophisticated, training necessary for their effective use gets longer. The information warrior must know not only programming but systems integration and systems theory, communications, electronic combat, security, artificial intelligence, logic in all its many forms (classical, fuzzy, and convergent), and statistical techniques. The information warrior must also know the customer's needs: the commander's intent, doctrine, and strategies. In addition, the information warrior should know something about specific media (land, sea, and space). Sending a college graduate to the field for a few tours of general expertise interspersed with training classes and then expecting first-rate information techniques in a more specialized tour later may not be adequate. The amount of information necessary to be an information warrior is immense, and the time required to master it will have to be at the expense of more general command instruction. If this tradeoff is to be made voluntarily, the results have to be rewarded commensurately. An integrated Information Corps with clear career paths and opportunities for command and success would do this.

As for ethos a divergence between operators and information warriors must be expected. Discipline under fire places a premium on certain qualities: courage, decisionmaking skill under pressure, good instincts, self-control, loyalty, and so forth. The information warrior, by contrast, must be highly intelligent, creative, independent, flexible, tenacious (to counter infamous 3 a.m. computer bugs) and maybe somewhat eccentric. The example of Admiral Grace Hopper will not excite a tank commander any more than General George Patton excites a bit twiddler. These qualities are not necessarily antithetical, and some qualities -- common sense, judgment, contrapuntal thinking, decisiveness -- are uniquely common to all warriors regardless of weapons. To seek such qualities in operators and not information warriors further relegates the latter to subordinate status.

Status, ethos, and training issues suggest the need for an Information Corps as well as a unified or specified information Command. Such a Command could produce unity of operation, advocates for change, and liaison, but it takes a Corps to provide doctrine, status, or continuity (e.g., information warriors who are evaluated by other information warriors).

Liaison:: In the same way that the information space of the various services converges, so too is the information space of the defense and commercial sectors. DOD uses commercial communications satellites and bought the bulk of Spot's imagery in the Gulf War; boaters use the DOD Global Positioning System. The defense and commercial sectors swap weather data; the DOD Global Grid is the military version of the National Information Infrastructure (which is a component of a global infrastructure). An Information Corps would play a major role in the development of a national information strategy and a complementary national military information strategy.

As the warning sign to builders "Call Miss Utility Before You Dig" suggests, both communities will have to shake hands before one or the other adds, subtracts, or alters its infrastructures. DOD used to formally liaise with AT&T when the latter was still dominant in telephony in the United States. Since then, the number of information players has multiplied -- and not just because AT&T has been rent asunder -- the influence of private networks has grown and number of various media has proliferated as well. In addition, as the DOD need for information intensifies, and its assets commingle with commercial systems, the volume of interaction will grow substantially. A common point of contact on the civilian side -- with its public and private players -- will never happen; a common point of contact on the military side is quite possible. A separate Information Corps would provide not only a common point of contact but common doctrine and outlook. With a national information strategy and a national military information strategy, human protocols would not have to be reestablished every time the two worlds come in contact.

Information Warfare

Just as the land, the sea, and then the air became realms of conflict -- and thus called out their own services -- so too might information be a realm of conflict, with similar implications. Information war, a clear official definition of which is yet to come, can take on several meanings.

If tanks fight tanks and subs fight subs why shouldn't information corpsmen fight each other? One increasingly popular concept calls for information superiority to be sought before seeking air superiority, which in turn, is a prerequisite to surface superiority. Two sides would duke it out to determine who could control communications on what frequency when, where and under what circumstances. Such conflict would feature jamming, deception, blinding, and firepower against key emitters, sensors, and other nodes. A related notion is for the data warriors to ascertain the other side's command-and-control architecture so that its weak points can be targeted. This would be coupled with the defense of one's own architecture either through a combination of engineering, bulwarking, massive redundancy, message prioritization, operational security, and deception. The relevance of this definition may be limited, though. Technologies such as frequency-hopping, spread- spectrum, and lobe control makes jamming relatively harder. Sophisticated network architectures leading to radical dispersal of command-and-control may complicate targeting efforts. However, it may take decades for our putative opponents to get such techniques right. In the interim, an information corps would be able to conduct such warfare more efficiently as an integrated team.

Another concept of information warfare posits it as the only stage of conflict. An information corps would be the body responsible for developing the doctrine and battle plan for such operations, then carrying them off. The Gulf clearly indicates that a ground force in conventional conflict cannot prevail against an enemy with air supremacy. Thus air supremacy alone may cause the loser to sue for peace prior to ground conflict. By the same token, information supremacy (if such a thing can be defined) may be sufficient harbinger of air supremacy and prompt a loser to sue for peace before a full- fledged air campaign gets underway. Defining such a hiatus before aerial conflict and the preceding information conflict (especially if the latter entails some destruction from the air) may need further refinement.

A related notion is information warfare not as a prelude to but as a substitute for conflict: e.g., non-lethal strategic warfare against key information systems such as air traffic control, space- based commercial communications, and financial networks. The advantages of such conflict for the United States, however, should not be too easily overestimated for three reasons: One, as the most sophisticated information economy, the U.S. is the most vulnerable to such warfare even if it is also most capable of conducting it. Two, networks can always be made relatively secure against attack either directly (better, more intrusive security regimes), or via backups that use completely different architectures--like land-line systems as backups to radio-based ones. Improved security costs more, but if network security is prerequisite to national security such price is more likely to be paid. Three, it may be far easier to isolate national information systems from international ones (either physically or via revoked permissions) than to make national systems crash. Such isolation, however, assumes that they and not we are hurt more by such barriers. For example, isolating a Chinese-dominated East Asia from the rest of the world may be akin to "fog in channel; Continent cut off."

Finally, the United States may provide information warfare capabilities as its sole or predominant contribution to an effort in which the actual fighting is done by such others as allies, or host nations. This, as noted earlier, has several advantages for the United States. It plays to our strength, minimizes casualties (to which we are becoming increasingly sensitive), eliminates most of the problem of lift (and its interdiction), and grants us at least some plausible deniability for the consequences. One of the strongest rationales for an Information Command is that such a campaign would come far more naturally to some future information CINC than it would to a regional CINC.

Functions of a Corps

Determining what an Information Corps does (on formation, its duties would be those of the units which comprise it) is tantamount to delineating the borders between the Corps and the services from which it would grow. The first concern is doctrine. The transformation of the Army Air Corps into the Air Force was more than a catch-all for those who flew planes; it was also an expression of a theory of war, to wit: the ability of airpower to transcend the ground situation and transform strategic conflict through aerial bombardment. The Marine Corps has its doctrine of amphibious warfare. Each service maintains its ability to comprehend war from its perspective.

An Information Corps would also have its doctrinal objective: to develop and exploit an integrated image of battlespace. This integrated image would, in turn, be divided and apportioned to meet the needs of various warfighters. Slicing and dicing would entail analysis, filtering, enhancement, correlation, data fusion, and whatever else is required to assist decisionmaking. The image, in turn, is an important component for decisions which range from strategy to weapons control. The bounds of such a system would vary from situation to situation. In some cases a coherent image would be used for centralized decisionmaking (such as an Air Tasking Order); in other cases the need for a better image would call forth efforts to collect further information (launching sensors). Some fire control solutions would be automatic, to take advantage of evanescent opportunities that a decisionmaking hierarchy would only slow down. Other images are background to on-the-spot decisions (tanks should not have to relay pictures of targets to a central mesh for a go-ahead before engaging them). Clearly the usefulness of a unified image depends on what percentage of the information involved in making a decision is generated by the shooter (coupled with what share of the processing necessary to transform data into decision is supplied by external algorithms). The doctrine is predicated on the assumption that nonlocal information (from other units or remote sensors) and analysis (from artificial intelligence) would rise in relative importance.

Should any operation that involves information, or alternatively command and control in its broadest context be part of a corps? This is probably too broad a sweep. Not only does everyone deal in one respect or another with information, but command and control tends to involve the top level of a hierarchy. To suggest that an Information Corps would become the top-level corps within DOD to which the services must report is presumptuous. To use such a corps to collect, process, transmit, and present information, then convey the resulting orders, however, is not.

At the very least, an Information Corps must encompass those elements which gather, assess, and distribute both silicon- and human-based information: an infosphere. Space would be a central component, since virtually every current use of space (e.g., surveillance, communications, navigation) is directly involved in information. Added to that would be chunks of the intelligence business, and the creation, operation, and maintenance of fixed-site command and control assets, information collection systems (such as ground-based radar and SOSUS), mapping, and meteorology, as well as the non-motile elements of the Mesh.

How far should an Information Corps extend into mobile information collection? Platforms as diverse as AWACS, JSTARS, AEGIS, P-3 squadrons, unmanned aerial vehicles, artillery trajectory indicators, portable radars, and the like are information-intensive and thus similar to fixed-site information systems; but not every function (like airplane driving) on such platforms is appropriate for the Information Corps. Consider the case of an AEGIS cruiser: it certainly collects a considerable volume of data, and much of it could be transformed into actionable targets for other platforms, but most of its functions call for other skills. Which among equipment maintainers, screen watchers, situation assessors, and communicators should be data corpsmen? Should they be permanently or temporarily assigned?

Such questions lead into difficult issues of scope, or how to prioritize among information flows. The forthcoming information architectures of the 1990s, complex as they seem, are relatively simple compared to what the twenty-first century's will bring. As popularly envisioned within the armed forces, the key function of today's information flows is to enhance the commander's situational awareness by developing an accurate, timely, and correctly detailed battlespace image. This image is devolved to the troops as per their needs. The number of sensors involved in this imagery is relatively small.

With the proliferation of sensors, the task shifts from providing the commander a view of the other side's tank columns, to providing operators a view of the other side's individual tanks. A smaller chunk of information goes to supporting command; more goes to supporting individual units of fire. The unitary view of what gets collected, how it gets collected, and what gets analyzed and presented therefore becomes much more complex.

Who determines this? The level of detail for the top commanders (who have only limited familiarity with the panoply of collection) is too complex for individual determination. The limited scope of sub-commanders (plus the fact that sensors will cross whatever artificial lines are established between them) makes their choice inappropriate. Letting the information commander choose may generate a coherent collection solution, but whence user input? The natural tendency may be to provide too much collection and information using the same logic that, in health care, is known as defensive medicine. These issues need considerable work.

A tougher question will involve the mix of military and civilians in an Information Corps. Should it be a Defense or Joint organization? Some functions of an information service can be best performed by military personnel with varying degrees of expertise and experience of the weapons systems with which they must interface. Other positions will have to be filled by computer jocks who are not disposed to military service.

Objections to a Corps

The difficulty in delineating an Information Corps suggests that creating one, at least in a platform-centric world, is at least somewhat problematic: it must interface with other command and control organizations, will remove critical functions of an operational unit, and may perhaps relieve some of the pressure of jointness.

Autonomy: Single-service cohorts are generally capable of operating autonomously in tactical environments, with little help needed from the others. Except as noted above, an Information Corps could not. If limited to fixed-site facilities, the Corps could at least function autonomously, but its value would depend on its ability to provide data to others -- it could complete few military missions on its own. But with dispersed sensors and emitters like UAVs, buoys, and listening posts gathering a larger share of the total data, a fixed-site Information Corps would be limited to strategic surveillance and distributed interactive simulations.

Including mobile elements in an Information Corps introduces command problems. Each unit of an Information Corps would have to report through its administrative chain of command, but it would also have to respond to the operational chain of command as well. Who, below the CINC or JTF commander, determines, for instance, when and where to deploy sensors? Who determines whether an aircraft is used for reconnaissance, electronic warfare, strike operations, or emitter dispersion? Do such needs respond to the requirements of the travelling unit (ship) or the deployed units of some information command (or under centralized control if not command)? The time required to resolve these issues or await their eclipse by circumstances (if ships disappear, shipboard problems do also) should not be underestimated.

A related objection is that even platforms whose exclusive mission today is to gather information may not necessarily retain that character. Reconsider UAV sensor packages. If the developers of this hardware and doctrine are information warriors rather than operators, they may not appreciate the potential of a UAV as a laser designator or a weapon rather than simply as a data collector. Such considerations have to be carefully melded into acquisition process.

Criticality: Every organization is an information organization; moreover, information is power. Removing information cadres from such an organization may promote several unintended consequences. Operational units may be tempted to duplicate their lost capabilities -- every important organization in the Federal Government maintains its own policy analysis shop. Besides wasting resources, it reintroduces the very coordination shortfalls an Information Corps was designed to overcome. Alternatively, affected military units may simply ignore the information they cannot control, relying on time- proven but obsolescent means to gathering information (reconnaissance in strength) rather than methods which technology makes more appropriate (sophisticated sensors). Thus, the very modernization that an Information Corps was meant to induce would be retarded by its formation. To avoid this shortcoming, strong leadership would be required inside and outside the corps.

Jointness: Finally, while creating an Information Corps may promote a joint battlespace image, it may retard other aspects of jointness. Removing the most important reason for the services to work together (they would instead liaise with an Information Corps) removes a large part of the impetus for operational units to work and meet across service lines. The need for joint deployment, joint operations, and, most important, joint thinking, remains, but the day-to-day practice of working jointly would be undercut by the act of shoving off certain joint duties to separate organizations. When the time came to act jointly, the various components would be far less prepared than if they had interacted on a day-to-day basis. The current concept of parallel jointness among peer services may need to be revised to accommodate a Corps that thinks itself superior working with operators who cannot help seeing the Corpsmen in their former support role.

Conclusions

When it comes to radical reorganization -- and forming an independent Information Corps certainly qualifies -- a first rule of thumb may be: when in doubt, don't. As wars are currently fought, the need for a data corps is, while perhaps inevitable, not necessarily urgent. Unlike, say, the Army Air Corps, which was a single identifiable operational arm, an Information Corps would have to be merged from several disparate organizations. By taking from all services, it would be opposed by all--a resistance difficult to overcome.

The logical conclusion, nevertheless, is that DOD should makes steps to form an Information Corps. The argument is that a corps would promote jointness where it is critically needed (information interoperability), elevate information as an element of war, develop an information warrior ethos and curriculum, and heighten DOD attention to the global civilian net. When threatened with the loss of personnel and resources, the services may respond that they are doing all of this and more. The greater the threat, the more meaningfully the services may respond. With luck, their response may address problems -- integration, doctrine, or ethos -- that would otherwise call for an Information Corps. Solving these problems, after all, was the original point. But can they do it as effectively as an Information Corps could?