U.S. Forces in Asia Map Out Consolidated Wide Area Network

May 2000
By Lt. Col. James G. O’Donnell, USA (Ret.)

Cost-effective scheme reduces threat to networks while increasing available bandwidth to thirsty users.

Communications specialists are proposing that the U.S. Forces–Korea change engineering and management approaches and follow the lead of commercial Internet service providers. The plan offers wide area network transmission bandwidth between the global defense information infrastructure and command, control, communications, computers and intelligence systems users on the Korean Peninsula, and it addresses several problems with the existing data network there.

Although survivability, reliability and restoration requirements are common to all communications systems, in Korea the threat from special operations forces and the proximity to North Korea bring a unique urgency to the situation. In addition, the demand for bandwidth poses a second challenge to the current transmission infrastructure. New command, control, communications, computers and intelligence (C4I) applications that transmit the large data files, video and imagery required for collaborative work between dispersed warfighters will overwhelm the present system.

These requirements led to the Korean communications infrastructure upgrade (KCIU), which established the U.S. military’s first large asynchronous transfer mode (ATM) network deployed in an overseas theater.

Finally, cost projections for expanding the current system of separate networks indicate an onerous requirement for new funding. Through using the shared bandwidth provided by ATM technology, the consolidated wide area network (CWAN) could result in substantial savings while making the emerging bandwidth requirements affordable.

The KCIU uses a U.S. government-owned-and-operated system known as the fiber optic communications system in Korea. It also employs OC-3, 155-megabyte-per-second microwave radios; and DS-3, 45-megabyte-per-second leases to provide a highly robust network that takes advantage of the self-healing nature of ATM technology. The system has significantly reduced communications outages during both day-to-day operations and critical exercises. The CWAN could extend this survivability to the remote user locations, significantly reducing the risk of communications failures during a war or contingency mission.

The existing system provides shared restorable bandwidth at the core of the communications system in Korea, resembling the base of an iceberg covered by several layers of networks. With the current architecture, operations and intelligence users all migrate their networks along the same technology lines: an ATM-based wide area network (WAN) feeding Internet protocol-based local area networks (LANs). However, because they are each building separate networks, the only shared and automatically restored bandwidth is at the KCIU core. A common WAN strengthens and extends the robust base of the infrastructure iceberg. The benefits include better survivability and more efficient bandwidth use.

The most critical factor in developing a survivable C4I infrastructure is eliminating single points of failure across a network. In the past, redundancy was built into each C4I system’s communications architecture. This was often accomplished by implementing alternate, dedicated point-to-point links between operating locations. In Korea, only the intelligence network has enlisted this concept to any extent. Technology advances now provide the ability to consolidate multiple physically diverse circuits between common operating locations into common user communications paths that all C4I systems can use.

The CWAN would present a more difficult target for special operations forces and a greater defense against the vagaries of accidents. Traffic does not stop at any network until all of the paths to the node are eliminated. With proper circuit setup, systems would continue passing data as parts of the transmissions paths are purged by whatever means. The approach also provides these benefits at significant savings over current network architectures.

A basic knowledge of the current architecture, CWAN physical connectivity and ATM network operation is key to understanding the bandwidth advantage provided by the CWAN architecture. Today’s system is based on time division multiplexing (TDM) technology. With this technology, each circuit uses dedicated bandwidth that is locked up by the particular system. This bandwidth is not available to any other system even if the communications terminals are turned off and no traffic is flowing into or out of the users’ facility.

The problem extends and affects the entire network, as it is an inherent characteristic of TDM technology. A TDM architecture limits each system to the bandwidth provided by its dedicated circuit, regardless of the traffic load presented by the terminal equipment. This presents problems to the network when large data demands are placed on it and results in long data transfer times and unacceptable operational impacts.

ATM allows sharing of the total node bandwidth, so that individual C4I systems would better support sending data bursts. The result is improved efficiency as each data network can take advantage of the total bandwidth provided to the node when it is available. When data traffic patterns are somewhat uncorrelated between the different systems, this characteristic allows the average effective bandwidth available to each system over time to be higher than in a TDM architecture. The physically diverse circuit paths are implemented as inverse multiplexed ATM trunks, which allows the T-1 circuits to be bundled into a virtual trunk of nxT-1, where n is the total number of T-1 paths into the local ATM switch. Each user then has access to nxT-1 of bandwidth from each node on an as-available basis. The user has access to a minimum bandwidth allocated to the network as well as additional bandwidth up to the total available to the node. This bandwidth on demand efficiently uses network resources.

Despite the survivability and bandwidth advantages of the CWAN architecture, it is not more expensive than TDM architectures. In January 1998, U.S. Forces–Korea and the Office of the Secretary of Defense for Command, Control, Communications and Intelligence co-sponsored a study titled “Combined Wide Area Network Implementation Proposal; Command, Control, Communications, Computers, and Intelligence Communications for the Future.” It includes an in-depth business analysis of the CWAN architecture. The results show that the CWAN would save a substantial amount of money over the course of five years.

The business analysis was performed using a cost effectiveness approach and supplemented by evaluating the relative strength of the potential benefits. It was based on the assumption that all systems in the study will require a minimum increase in bandwidth from T-1 to DS-3 by fiscal year 2002. Although this appears to be a large increase in bandwidth, many user requirements currently exceed this level. Already some users are identifying OC-3 requirements in the near term.

The analysis compared the CWAN with future stovepipe C4I system costs in Korea. These were estimated across three functional areas: record communications, data exchange and video broadcast/teleconferencing. The study looked at the period of fiscal years 2000 through 2004. Total operating costs for the CWAN are estimated as $48.9 million over five years and include a first-year capital outlay of approximately $4.5 million. Additionally, the cost of the contractors needed to operate the system and recurring costs such as leased communications paths and maintenance of the equipment were incorporated.

The study findings indicate projected future costs for the current C4I stovepipe systems were higher for both life-cycle costs and net present value. Initial analysis shows that an approach using the CWAN would yield a 67 percent reduction in life-cycle costs with a comparable savings of 64.5 percent in net present value over stovepipe systems.

A CWAN is now possible because of significant advances that support advanced network switching and encryption. The availability of these technologies in the near term will permit the implementation of a true common-user communications infrastructure, one that can provide a more robust, survivable capability while reducing future costs to realistic levels. It calls for a common ATM transport infrastructure providing connectivity from the communications technical control facilities down to the user-facility level. Individual C4I systems managers would no longer have to perform a wide area communications function. Instead, point of presence ATM switches would be installed in U.S. Forces–Korea facilities so that individual C4I units would only be responsible for connecting their systems to the local point of presence to enter the ATM infrastructure.

The engineering and management approach for the CWAN is based on the business practices of commercial Internet service providers (ISPs). For example, ISPs typically install an Internet point of presence that is made available to individual users. When a user connects to the point of presence, the ISP directs and controls all phases of the connection from the individual router device out to the network and ultimately to the end destination. The ISP controls the Internet protocol and the addressing schema, and it dictates the router WAN access port configuration. It also manages, maintains and controls every facet of user WAN connectivity. The CWAN would include a centralized operations support activity that manages CWAN point of presence devices, encryption devices, network addressing and connection plans, procedures and policy proposals. This centralized operations support activity would be the ISP for U.S. Forces–Korea.

Implementing a CWAN would allow these forces to take advantage of technology advances, consolidate its many wide area networks, streamline network management and improve user satisfaction. The infrastructure would ultimately serve more than 140 user facilities at 50 locations on the Korean peninsula. It would significantly enhance the survivability and adaptability of the Korean theater communications network infrastructure.



Lt. Col. James G. O’Donnell, USA (Ret.), recently retired as chief, communications architecture and plans branch of U.S. Forces–Korea, J-6.