«DEFENSE SCIENCE BOARD SUMMER STUDY TASK FORCE ON INFORMATION ARCHITECTURE FOR THE BATTLEFIELD DTlC OCTOBER 1994 S ELECTE APR I 0 1995' G i 95-01137 I ...»
The likely continuing practice of drawing partial portions of forces from standing assets, and the need for flexible inc±,.ase or decrease in their sizes as a Joint Task Force executes its mission, demands that "scalability" be a key attribute of the supporting C41 systems. Our C41
systems should therefore have certain attributes:
"Theymust be dwigned - or incrementally upgraded (in the case of legacy systems) to support reconfigurability options. These should allow the warfighter to make or change the interconnections among systems during the course of the missions.
" They should provide information management options among the C4I systems in the battlefield. These options should accommodate changing needs for access and changing needs for protecting information.
While it would be preferable to allow the warrior to tailor the information management needs to the situation on the battlefield, today's sensors, communications, and ADP systems cannot accommodate much reconfiguration. There was a need to connect information from overhead assets to Patriot Batteries during Desert Storm. However, since the systems weren't designed to support that link, a relatively complex set of connections had to be established onthe-fly. Had there been disruption attempts made on the patched communications, field commanders would have had few options for recovery. If a commander deploys four aircraft on a mission, he doesn't need the same communications and information resources that would be n.eededfor two wings going into a theater. Depending on the size of forces, fusion points, communications strategies, etc., a commander will need considerably different support systems and information management strategies. These and other examples suggest that there are some fundamental reconfiguration properties that should be considered for the warfighter's quiver.
Access management would allow the warrior to select how, where, and when data and information are fused and disseminated. Part of access management would be the metering of information to prevent overloads; however, the ability of commanders to access management flow. Consequently, assimilate information influences the speed and volume of information includes many interrelated parameters.
I MC10n' Vulnerability management is similarly complex. Detection of failures, failure mechanisms, recovery processes, aad the management of risks due to information disclosure yv. the risk of failure to provide needed information rapidly must be addressed.
The properties of the communications systems must be considered. Some lend themselves to broadcast or publishing dissemination strategies, whereas others perform better as direct point-to-point links. The type of information being exchanged needs to be matched to the media options available. Those who reside on fiber optics will have information flow at teraflop levels while the tactical, mobile, satellite users will be at a 10,000-to-i bandwidth disadvantage. Technology pursuits must take this variance into consideration.
2.4 Dynamic Information Management Battlefield decisionmakers are at risk of being inundated by data when they need useful information to build knowledge. Decisionmakers receive information effectively in different forms. Filtering, fusing, and correlating data to selectively provide information to decisionmakers needs to be emphasized in our migration of systems. Modeling, simulation, knowledge mining, and human factors disciplines need to be involved to improve selective information dissemination to decisionmakers.
The warrior should have dynamic control over the information form and flow. He should be able to lay out information needs tailored to the particular situation. For each type of information (e.g., air surveillance, imagery, friendly force status, etc.) he should be able to specify what information he needs, in what detail, updated at what frequency, with which access controls, fused with which other information, displayed in what form. One might imagine the commander conceptually filling out the chart in Figure D-3. For each type of information that will be circulated around the battlefield, the commander is asked to indicate where the information should flow, the detail to be provided, the response time for delivering information, etc.
Within the constraints of the current situation, his information officer would then "reprogram" the sensors, communications and ADP to respond to these needs. This scenario is not possible today. The systems are not capable of being rapidly "reprogrammed" and staff do not have the technical capability or tools to do the job. This is an important refocus area for R&D investment.
Point-to-point communications are dominant today in the distribution of information for the battlefield. Voice circuits, message traffic circuits and remote computer connections all play a part in achieving information distribution. While this permits the greatest degree of information customization, it is very costly in terms of communications resource utilization.
Threat Alerts Artillery Locations Figure D-3 As is discussed in Section 3.8 of the main report, broadcasting (publishing) could be used to off-load a notable fraction of the information distribution workload, without adverse effects on quality of the information. For example, certain status of forces information, environmental information, and GPS time are very well suited for broadcasting. Some broadcasting is used today, but only through custom data links such as JTIDS and TRAP.
Commercial broadcasting can open the range of these kinds of services.
To allow wider distribution of information, it is most important that information "receipt mechanisms" be low cost. Low cost is achieved through a combination of the design of the overall system architecture, technology advances and high-volume automated (commercial) production. The low-cost GPS receiver is an excellent example of all three of these factors. It is reasonable to have high-cost TV studios, expensive broadcast facilities and costly satellite relays as long as the consumer's TV set is cheap. If the receipt mechanisms are in the million dollar price range, we can be assured that the product will never be distributed beyond a privileged few.
If we are to significantly increase the flow of information to military users, we must also
add several architectural elements that do not exist in the structure today:
"* The information needs to be packaged into readily usable forms. Contrast the typical long military message (ALL CAPS and annoying Headers) with a Time Magazine-like format. To do this requires that a creative, professional, and conscientious editing function be added to the distribution process.
"* Methods, standards, and development of tools are needed to monitor and assess data quality. Most military databases today have no formal procedures for quality control of the content. For those that do have procedures, their standards vary greatly and those standards are generally unknown to the user. It really helps to know whether you are reading an article from the National Enquirer or the Washington Post. (It is left to the reader to make a judgment on the relative quality control.)
Today's systems for distribution of information can be enhanced, but new approaches and mindsets are needed to do it effectively. Request and delivery of information through the multi-layered information system could be substantially augmented by broadcast systems and direct database access arrangements.
In order to maximize effectiveness, an analysis of information distribution alternatives is necessary, using a variety of communication media. New commercial technology may provide added capacity and lower cost user equipment. Of course, potential vulnerabilities associated with commercially-based concepts would need to be accounted for in any management decision.
3.0 THE NEED FOR A JOINT ENTERPRISE ARCHITECTURES FRAMEWORK
3.1 Convergence 3m. Divergence to Joint C41 For the Warrior
FigureD-4 There are several ongoing programs devoted to improving C4I capabilities. Each of the Services and Agencies has programs, devoted to battlefield support, which are attempting to adhere to an architecture defined for promoting interoperability. As is indicated by the curving arrows in Figure D-4, the programs are paying some attention to the need to migrate into a I ~unified C4I structure by conforming to the GCCS migration plan. However, processes are ! IU needed to ensure individual programs have adequate cost and schedule provisions to allow the separate initiatives to achieve effective interoperability and a common operating environment. Until a process is put in place to ensure the joint warfighter's requirements are strongly considered, the well-intentioned but unique Service and agency programs will tend to drift away from migration objectives.
Current acquisition practices exacerbate the tendency to drift. Since each program is independently supported by mostly independent agencies, a joint corporate perspective is not D-8 built into the acquisition process. The warfighting CINCs and JTF commanders have little influence on systems under development or being modified, but they have perhaps the most at stake when systems reach their ultimate application. The joint warfighters' concerns should be represented during the acquisition process to ensure the C4I systems that will support the warfighter have maintained pace with commercially available technology and will intermesh well with legacy systems.
Legacy systems must either be migrated into or interfaced with common systems. The motivation to diverge from a common joint interoperation structure is aggravated by the need to maintain compatibility with Service-unique legacy systems which are not targeted for migration.
There is a need for establishing a process, in a manner akin to that used for the Internet, which identifies incremental improvements and ensures each can be accommodated and accepted by the other participants. The part of the Internet process which establishes standards by consensus, allows continuous integration of improvements, migration of standards, adaptation of commercial products, and distribution of value added products has been shown successful. Some variant of that process is appropriate to institute for the DoD. Unlike the Internet, the DoD will need a method of measuring overall cost and benefit of modifications, and ensuring appropriate benefits accommodate each incremental change. This requires refocused investment to develop and/or acquire tools to facilitate these efforts.
The process should include provisions for accommodating the limitations of legacy systems and easing their transition to modernization. This process should be recognized as a continuous process; there will always be a need to manage transition from old to new systems.
To provide the developers with the opportunity to purchase the latest, most cost effective components, enabling standards should be used. Only where absolutely necessary as a part of our migration strategy should mandatory standards be applied. In the past, each DoD component has developed information systems architectures in its own way. Standard definitions of architectures and architectural objects are missing. Standard interfaces are also essential. There is no commonly accepted joint taxonomy of information systems architectures.
Airlines specify needed products for aircraft at the "box" level (inputs, outputs, form factor, reliability, speed, etc.) - not in how the box does its job. DoD should consider such an approach (ARINC approach) for information systems in DoD. This could potentially reduce the large number of standards that DoD maintains for information systems. We recommend DoD investigate the feasibility of simplifying standards for DoD information systems by specifying at the "object" level rather than at the "how to" level.
Architecture Principles 3.2 Key to applying R&D resources to the improvement process, is an understanding of "architectures," and how that understanding can facilitate the investment decision-making process. Unfortunately "architecture" is almost a Tower of Babel when it comes to definitions.
The word "architecture" is best used in the form of an adjective (architectural style, architectural feature, architectural standard, architectural description, etc.). However, it continues to be used as a noun; and, in that form, promotes much ambiguity. Nevertheless, a generally accepted concept is that when something conforms to an "architecture", it has some underlying order or structure. Further, while in one sense "conformance" to something D-9 implies restrictions or diminished flexibility, in the sense used here "conformance" provides order or structure which has some significant benefits - including enhanced flexibility.
In current DoD usage, one benefit is in efficiency of resource use. If "entities" are computer systems, fighting forces, weapon systems, etc., there can be many options for interconnecting them if they conform to various architectural standards. If components within a system adhere to architecture rules which minimize interdependence among components, a good architecture will offer, as a second benefit, the ability to efficiently modify a system by improving components or replacing them with newer components. Lastly, to the extent these arcuitectural principles are developed in the civil marketplace, there will be many conforming components available for the DoD to select for new systems. This latter situation provides benefits in both cost and schedule.
For the warfighter, the "architecture" theme can mean better interoperability, changeability, and quicker, cheaper capability in the field. While the concept and objectives are relatively simple to understand, achieving the benefits requires both a more specific definition and a more explicit process for defining and preserving architectures.