Military grapples with missing links to its global network
Enabling network-centric operations will require a broad set of standardized policies to ensure that systems interoperate and that the proper policies accompany the information as it is handed off from one domain to another.
The Defense Department’s commitment in 2002 to the concept of network-centric warfare made the idea of full network integration — from headquarters to the warfighter in the field — a hot topic in government and industry. But the idea that those on the front lines should have access to a broad range of information to make decisions on the spot is not a new one.
In this report:
DOD maps out infrastructure to support network-centric force
“Net-centricity has existed as long as there have been more than three people on the battlefield,” said Terry Morgan, director of business development for Cisco Systems’ Global Government Solutions group.
In its current form, the concept of network-centric operations entails optimizing networks for the rapid, trusted point-to-point exchange of information.
“It isn’t really about warfare,” Morgan said. “It really covers operations of all types.” Warfare is just another operation.
In its simplest definition, network-centric warfare, network-centric operations or network centricity is the ability to enable persons anywhere — on the battlefield, say — to pull the information they need when they need it, rather than waiting to have the information pushed to them from a central source. But net centricity is not simple to achieve. The DOD’s principal networking backbone, the Global Information Grid, or GIG, is a distributed, stovepiped system of networks with varying degrees of interoperability that hamper the ability to achieve net centricity, according to the department’s architectural vision for a new net-centric GIG.
“In addition, the current GIG is static rather than dynamic; it cannot quickly adapt to satisfy unanticipated needs and users,” DOD’s vision statement says. “Most importantly, the current GIG is not suited to support [network-centric operations] — it does not support the ability of warfighters and business and intelligence operators to leverage the power of information.”
To support the network-centric operations, the GIG must evolve to support the exchange and management of information among all users through well-defined interfaces, according to the architectural vision: “A key element of the future GIG will be its ability to extend that visibility, accessibility and sharing to unanticipated users.”
The ubiquity of both wired and wireless, fixed and mobile networks today and the convergence of multiple protocols and services onto a common IP infrastructure, now make possible this network-centric GIG, said Bob Natale, chief engineer at Mitre.
“This provides the capability of realizing network centricity rather than just having it as a vision,” Natale said.
But although the capability is available, DOD is not yet at the point of having full network centricity, he said. Bandwidth is necessary for network centricity, but it alone is not sufficient. “On the network operations front, a lot has to happen.”
The ability to transport multiple protocols and services across a common infrastructure does not ensure that all elements of that infrastructure are truly interoperable or that network management and policies are compatible. That means DOD cannot ensure that the information being transported is being handled in the same way at all stages of its journey.
“As you make a handoff from one administrative boundary to another, there needs to be a common way to recognize that data, and carry things like quality of service information and other policy information,” said Tim LeMaster, director of systems engineering at Juniper Networks. “There has to be a common set of standards for how that handoff is established.”
Those handoffs currently are handled under “gentlemen’s agreements” between carriers on how quality of service and service-level agreements are honored by third parties.
“Generally speaking, there has been very good cooperation between carriers,” LeMaster said. “But there are no standards for that.”
Putting formal standards in place could help leverage existing network infrastructures to deliver a growing array of services on an ad hoc basis, which could in turn offer great tactical and financial advantages, depending on what sector you are working in.
“The trick is that there are going to be nodes all over the world,” said Juniper’s Todd Shimizu. “If I am going to pull it off, there needs to some sort of federated framework to enable management of quality, security and [in the case of commercial content] remuneration.”
Shimizu is Juniper’s program manager for IPsphere, an industry group that now is part of the TM Forum and is working on the development of such a framework. The TM Forum is focused on standardizing network management among a growing variety of networks and network operators that now includes mobile, wireless and satellite, as well as fixed networks; and content providers and enterprises such as DOD as well as traditional carriers.
Setting standards
Standardization makes good economic sense while enabling delivery of network centric services to DOD and other defense enterprises, said TM Forum president Martin Creaner.
“There is a lot of off-the-shelf equipment that, if there is an adherence to standards, is reusable in the defense industry,” Creaner said. Standards can help reduce or eliminate what he called the “integration tax,” or the cost of making things work. “Integration tax is a huge overhead for any organization. For every one dollar you spend on software, you spend another five dollars to make it work in your environment,” he said.
IPsphere and the TM Forum are not alone in the quest for network-centric standards. Cisco’s Morgan is chairman of the Network-Centric Operations Industry Consortium, an industry group developing specifications for fully-integrated networking.
NCOIC is a not-for-profit association of more than 75 member organizations from 18 countries, established in 2004 to advance deployment of net-centric applications. Members represent defense contractors, systems integrators and IT companies along with government agencies and emergency responders. NCOIC advisory bodies include government officials, standards groups and other public- and private-sector stakeholders. The members came together to chart a development path toward network centricity, including specifications for interoperable equipment and architectural guidance. It also has produced a tool for analyzing the degree of network centricity in a system.
The analysis tool was developed by Lockheed Martin and donated to the consortium. It is a methodology for questioning the ability of a system to hand off to and to accept data from other systems while maintaining policy and management information that is necessary to enable network centricity. The capabilities and capacities of interoperating systems need to be matched, or at least understood, in order to seamlessly exchange data.
Network centricity is not just for DOD any more, said Kevin Orr, director of Cisco’s federal defense unit and a participant in NCOIC.
“As the trend matures, we are seeing across government a movement toward off-the-shelf technology” and the adoption of network-centric goals for operations, Orr said. “It’s the same principle over again. It comes down to command and control.”
One agency looking to network centricity to enhance a vital command and control structure is the Federal Aviation Administration, which recently entered into an agreement with NCOIC to help in development of FAA's national airspace transformation program, called the Next Generation Air Transportation System.
FAA plans to phase in the system over the next 16 years, replacing current radar technology with a more accurate and efficient satellite-based system. At the heart of the new infrastructure is the Automatic Dependent Surveillance-Broadcast system, which lets airplanes determine their position using a global navigation satellite system and broadcast that information to other aircraft and ground stations, rather than depending on ground-based radar.
But industry observers have complained the agency is not moving quickly enough to meet its 2025 goal of implementing NextGen, and the Government Accountability Office says FAA needs to provide incentives for the airline industry to invest in the equipment needed to make it work.
Under a five-year, $10 million contract, NCOIC will analyze and evaluate NextGen's enterprise architecture views, products, plans, net-centric patterns and operational concepts. It also will provide an industry voice in the form of recommendations for using net-centric standards requirements in NextGen procurements to enhance interoperability.
Incorporating an industry review of the program this early in the development process is a new concept that FAA hopes will help speed system development, reduce cost to FAA and encourage broad airline industry adoption of the new technology.
Building networks
The IPsphere project is developing sets of interface specifications and standards to enable network-centric applications. A Defense Interest Group within TM Forum is working with the IPsphere project to develop a common framework for specifying and delivering services. Within the defense community, “there is a consensus that they don’t want to be in the networking business,” Shimizu said. “They want to be able to publish their needs in a template that service providers will be able to use to build the services.”
IPsphere’s concept of a service script could enable this, allowing end users to pick and choose providers that meet their needs with standardized services.
IPsphere was formed in 2003 by Juniper and a number of large network carriers looking for answers to the challenge of managing networks and data in multi-domain environments. It became the IPsphere Forum to pursue standards development in 2005 and became a project of the broader TM Forum in 2008.
TM Forum was established in 1988 as a telecommunications industry group focused on management of large carrier networks. Its scope has expanded and today it claims more than 750 members, including tier 1, 2 and 3 carriers, fixed and mobile network providers, content providers, hardware and software vendors and systems integrators.
The IPsphere framework is a service-oriented architecture, Shimizu said. It enables the abstraction of objects that can be mixed and matched by service providers according to the needs of users.
“You only publish the things you want to do,” and the service provider uses the objects to set up the business rules for them, Shimizu said. “The Internet is pretty light on rules and control.”
The first set of specs in the IPsphere framework containing the basic functional blocks were released in June 2007, followed by a second set covering session services and resource management in February 2009. Initial specifications were field tested by an international group of telecom carriers during last year’s TM Forum gathering in Orlando, Fla., to dynamically set up end-to-end video services. They were further tested at this year’s TM Forum gathering in Nice, France, to run a more complex high-definition video conferencing application.
The next release of specifications is expected to be completed late this year or in 2010 and will update the original functional blocks with a greater level of detail.
Rules of order
Better rules also are needed for handling information within the enterprise as well as between networks to fully enable network centricity, said William Chang, a technology fellow at Booz Allen Hamilton. The Defense Information Systems Agency is moving to a policy-based enterprise level of management that is based on the behavior of information resources on the network, he said.
“Network centricity is not about the network or networking,” Chang said. “It is about information sharing and information management.”
With all of the attention being given to network-centric operations, are we network centric yet?
“It’s an evolving situation,” with progress being made on both the technology and policy fronts, Mitre’s Natale said. He said he expects to see additional progress over the next couple of years. “But it is not a global reality. There is a lot of work to be done.”
Within the defense community, the concept of network centricity itself might change as the new administration and new departmental leadership gain experience in fighting new kinds of wars. The need for the warfighter on the ground to be able to pull essential information into the field is real, but the military is inherently leader-centric, Natale said. The idea of complete network centricity might have to be adapted to take into account the need for a commander to make decisions at some point.
“You might see some rethinking,” Natale said.
“There is still a lot of work ahead,” Shimizu agreed.
Specifications and standards still need to be developed and refined, and buy-in across the complete spectrum of stakeholders still is needed, he said. Discussions are ongoing between industry and the defense community to apply the network-centric framework for service provisioning and management to the new model of military operations.
“That is a priority for all sides,” Shimizu said.