Preparing for a Network-Centric World
Research seeks to speed military data integration through advanced modeling, new computing techniques and policies.
Future warfighters will benefit from enhanced information technology infrastructure and hardware as well as support services being developed by the U.S. Defense Department. As a part of its ongoing transformation efforts, the military is changing established cultural and business practices to meet the computing needs of widely dispersed U.S. forces.
Network-centric warfare methodology is beginning to extend beyond sensors and weapons systems. By linking communications and logistics support networks across traditional service boundaries, the Defense Department is creating a fully integrated system for sharing critical operational data among commanders.
The Defense Information Systems Agency (DISA), Arlington, Virginia, is responsible for managing and deploying the U.S. military’s command and control infrastructure. According to Dawn C. Meyerriecks, DISA’s chief technology officer, the organization is undertaking near- and long-term goals to change its culture and business services while developing new capabilities. The agency is moving away from what she refers to as “boxology” and toward a model based on managed services. DISA not only will continue to provide information technology infrastructure and hardware support for warfighters, but also will develop methods to reduce the services’ maintenance and management responsibilities. She explains that this will permit the services to concentrate on their mission by deploying troops and material, instead of worrying about how much fiber optic wire to lay across a theater of operations.
The new strategy represents a conceptual challenge because its underlying business model has not worked well in the commercial world. Firms such as Cable and Wireless attempted to provide managed service packages, but they have since exited this market. Two firms that have approached some success are America Online and Microsoft Networks, but their systems are not considered viable for replication in a mission-critical application, she observes. DISA is trying to accomplish this goal itself.
By crossing the line between traditional hardware and infrastructure support and software applications services, DISA is rethinking the ways it interacts with the services’ logistics chains. Meyerriecks explains that traditionally, DISA provided equipment to the military—servers, computers and other hardware—and provided support for them once they were in theater. One issue was cultural. When a failure occurred, warfighters wanted someone to be responsible for the mistake. “Information technology was one of those areas where we actually gave the services the boxes, and if they broke, officers could go find a private and chew him out as opposed to working the chain of command to get someone who can actually work things,” she says.
But growing net-centricity introduces more variables into systems that are beyond the control—or ultimate responsibility—of on-site support personnel. DISA is beginning to change this attitude in part because the services no longer have the resources to pile on more equipment when they require a new capability. “From a technical perspective, it comes down to architecting—and by that I mean really designing how the boxes and services have to string together so that you can give an appropriate level of service in a distributed environment,” she offers.
Another issue is that networks tend to be managed by different groups of specialists. “When something goes down, it’s really hard to put Humpty Dumpty back together again and figure out exactly where the problem is that’s keeping the commander from being able to access a data source because there are so many players involved,” she says. To change the current methods, which are mainframe and applications services oriented, an enterprise systems management strategy covering the entire network from end to end, irrespective of organizational boundaries, must be put into effect.
Other issues involve how the services and DISA interact in a business sense. For example, Meyerriecks notes that one of the services discovered a data source that was of high value to the Defense Department. However, because DISA charges for central processing unit (CPU) and storage requirements, when the data source was heavily used, that service paid a high computing bill. “Not only were they paying for their use, but also for everybody else who was accessing their data and causing their CPU utilization and storage requirements to go up,” she says.
Process and strategy changes are all a part of DISA’s reorganization (see page 29). Because the agency was not organized properly for net-centricity or to meet the Defense Department’s transformation goals, DISA Director Lt. Gen. Harry D. Raduege Jr., USAF, initiated a complete restructure, Meyerriecks says. The changes went across the organization, from personnel activities and services to funding and billing.
The reorganization also seeks to change DISA’s culture. She notes that though the agency had gone through reorganizations in the past, it simply rearranged different components without modifying the way it did business. This comprehensive approach was important because DISA had been pushing for such changes in the other branches of the Defense Department. “It was ‘physician, heal thyself.’ We’ve been saying it to the services, but we hadn’t really eaten it, at least from my perspective, on our agency side,” she says.
DISA created task groups that can be assembled to deal with a crisis and then split up to form other groups to meet new needs. The teams grew out of suggestions DISA was proposing to the services. Agency staffers now can expect to move around within the organization during the course of their careers. “You are assigned to this team for now, and when that job is over, you’re still going to have a job, but you’re not going to be at the same desk for the rest of your professional career,” she says. As a result of these efforts, the agency is more flexible and capable of reacting to sudden changes in service requirements.
Through such initiatives, the agency is trying to put its business activities on a more market-driven footing without sacrificing its strategic perspective. By taking lessons from industry and applying them to the defense world, the reorganization has sought to maintain the Defense Department’s operational rules, but be more flexible and agile in terms of responding to situations.
New technologies also will impact future DISA operations. One major trend that will affect DISA and the Defense Department within the next five years is heterogeneous grid computing. Meyerriecks notes that, for this type of networking to be successful, it must be supported by a world-class Web service. For example, while Web-based services are available, the security model for them is still a work in progress. “We can’t employ Web services unless we can figure out how to make them secure and have a pedigree,” she says.
Other related issues involve providing security while supporting dynamic communities of interest across different areas such as homeland security, homeland defense and coalition activities. Converging them so they can securely share data and supporting the various services and applications are problems that have yet to be solved, she explains.
DISA released a policy for wireless data systems that provides broad engineering and program guidelines within a flexible policy structure. It permits the services to apply wireless systems according to their needs without compromising security or mission capability. The agency helped rewrite policy regarding wireless applications such as instant messaging and devices such as Blackberries. This new policy supersedes older guidelines that were too narrow in scope and binding regarding permissible systems and applications. Meyerriecks observes that the revised rules attempt to strike a balance between letting the services do what they must to accomplish their mission and providing enough guidance so that the Defense Department’s communications infrastructure is not compromised.
The new rule also outlines security for wireless systems. Meyerriecks explains that robust security for wireless systems is available, but it is proprietary. Because the services are going to use the technology and apply the security they trust to best support missions, DISA is working with vendors to create specific form factors. These guidelines eliminate the need to write a different interface for every tool that operates in the infrastructure. By working at the application level, DISA is creating a wireless framework that incorporates industry standards, including Web services and identity management. “Those aren’t particularly wireless things, but they come as the packaging—part and parcel of the overall technological revolution that’s happening,” she says.
In addition, DISA is working with industry and research groups to develop new modeling and simulation techniques. The goal is to escape the testing-to-exhaustion methodology that characterizes every instance that a software application will operate under while finding a safe range where it will work predictably. “We still haven’t figured out how to characterize the environment in which it’s safe to use something that’s less than perfect,” she says.
IBM is working with DISA to develop predictive analysis and modeling techniques that will help commanders make decisions about which information technology and software components to deploy. DISA engineers are using a data model in a distributed development environment in their efforts to design this flexible system.
This modeling vision extends beyond the current five-year horizon, Meyerriecks explains. She is sanguine about being able to develop individual pieces of the framework. She notes that, by using grid computing techniques, future systems operators and warfighters may be able to obtain a picture or map of the operational environment. Such a map would also track points of need, including multiple requests for specific kinds of intelligence data, from a specific point in the grid.
The agency is examining how discrete groups, such as human and animal populations and biological organisms, behave under certain conditions. The same science will then be applied to information technology systems because computer networks are becoming so complex that they are beginning to resemble social and biological systems, she explains.
Although such a modeling capability will probably not be available for at least eight or 10 years, it may permit future commanders to assess the operational capabilities of their systems and components accurately. Meyerriecks cites the example of a diving watch. It can be used above sea level and is qualified to a certain depth before it fails, providing the user with an operational range. “We have to characterize software in the same way. We’ve never done that—we’ve always said, ‘Here are all the instances that you can invoke this module, and it will behave the way it needs to,’” she says.
Citing a study done with the U.S. Navy, Meyerriecks notes that the services have built what she describes as “beautiful integrated stovepipes that presumed that we always knew what the threat was going to be.” As recent combat operations have proven, enemy forces are no longer likely to operate under predictable Cold War patterns or follow rules of engagement. The exercise was a missile defense scenario. Under traditional stovepiped systems, it was impossible to include radar information that was not intrinsic to a ship’s system. However, by allowing a complete interconnection between different sensors and platforms in the scenario—such as U.S. Army Patriot missile batteries—the Navy was able to engage 100 percent of all the incoming targets and increase kills by 40 percent. “That’s the impetus for this net-centricity—characterizing components so you can dynamically compose a response. This differs from going back to the software factory and asking them to put together a new work flow with different rules of engagement because the bad guys have changed the rules,” she says.
DISA is experimenting with a number of different technology approaches. One of the agency’s efforts is to make computer network architecture meaningful to both systems engineers and warfighters. Meyerriecks notes that this does not involve training personnel in the intimate details of an information technology framework. The goal is to expose both engineers and warfighters to a methodology where technicians can make decisions based on an architecture, and soldiers can operate based on capabilities derived from what is available in that architecture.
The ultimate goal of this work is to make information technology ubiquitous and invisible. This is the next major challenge facing DISA. Using the telephone system as an example, Meyerriecks notes that civilian users do not need to know where the switching station nearest to their house is located, but warfighters must rely on communications support maps outlining where every station is to access information technology services. The goal is to make the technology and linkages completely invisible. “As hard as integration is today, we ain’t seen nothing yet,” she says.
Additional information on the Defense Information Systems Agency is available on the World Wide Web at http://www.disa.mil.