Disruption-Tolerant Networking Research at BBN

Survivable Policy-Influenced Networking:
Disruption-tolerance through Learning and Evolution II
(SPINDLE II)

The US military is transforming into an agile distributed network-centric force. The new doctrine is critically dependent on access to mission information even under temporary disruptions to connectivity in the Global Information Grid (GIG). DARPA's Disruption-Tolerant Networking (DTN) program is developing technologies that enable access to information when stable end-to-end paths do not exist and infrastructure access cannot be assured. DTN technology makes use of persistence within network nodes, along with the opportunistic use of mobility, to overcome disruptions to connectivity.

Traditional TCP/IP networks rely on stable end-to-end connectivity -- an identifiable path all the way to the destination. In the Department of Defense's wireless tactical networks, connectivity is often disrupted by terrain, weather, jamming, movement, or destruction of nodes. This disruption to connectivity makes it impossible to predetermine a path, which halts the flow of message traffic.

The current Internet model of locator-based access relies on strong connectivity to naming, caching, and search infrastructures to deliver data successfully. In contrast, the system BBN is building will provide disruption-tolerant content-based access to information. For example, if a user wants to see a map of Baghdad that is no longer available via a network connection to a map server, his request could be satisfied by an accessible node that had previously stored the map.

In contrast to TCP connections, DTNs communicate opportunistically using episodically or intermittently available links. In phase 1 of the project, BBN successfully organized information into bundles (a concept developed by the DTNRG, an Internet Research Taskforce Working Group) rather than packets. Bundles are routed through "custodians" that augment the capabilities of traditional routers by persistently storing the bundles and then advancing them to the next available node en route to their destinations. This opportunistic routing may require novel methods of advancing the information, such as using unmanned aerial vehicles to carry message traffic when there is an obstacle in the path -- geographic or structural -- or in the presence of an enemy threat.

BBN demonstrated key performance goals using DTN technology: 100%-reliable delivery of data with less than 20% availability of links with greater than 80% utilization of link capacities. The DTN approach consistently outperformed traditional end-to-end approaches across a wide range of network disruption. Under certain worst-case network dynamics, DTN was able to deliver data reliably, while the traditional end-to-end approach broke down and delivered no data at all. However, to fight the wars of the 21st century, DoD requires networks that can provide disruption-tolerant access to information. Continued research and development beyond the phase 1 demonstrations are required for phase 2 in order to achieve a transitionable and evolvable, long-lived capability for DoD.

In support of this goal, BBN is building a DTN bundle-message processor and related technology.

Key contributions from the SPINDLE II project are:

Using a prototype system in phase 1, we clearly demonstrated the benefits of DTN technology. To support the program's phase 2 objectives, further research and development are needed in the areas of scalable DTN routing, policy-based resource management, late binding of name attributes, and security management of persistent information within the network.

Building on phase 1, our enhanced vision for follow-on phases of DTN requires a radical departure from the current Internet model of locator-based access to information. The traditional model relies on strong connectivity to naming and search infrastructure, but today a new model is necessary, one that provides disruption-tolerant content-based access to information. When a warfighter needs that map and it happens to be cached on a computer in the team he is connected to, his query must be able to get the information even though he is disconnected from its usual source. To realize this enhanced vision, we are developing distributed, opportunistic algorithms for caching, indexing, and retrieval in a manner that maximizes the availability of information required for the mission, even when disconnected from the Internet.

To guide further research and development, it is also vital to gain operational experience with the technology at this stage. To that end, we develop prototype-deployable DTN systems (hardware and software) and make them available to selected DoD transition partners to enable evaluation of DTN within the context of the GIG. Using an open-source, standards-based core with a plugin architecture and well-specified interfaces, we enable independent development and insertion of innovative DoD-relevant technology while allowing the core system to be refined and engineered within a COTS context.

Presentations

Documents and Reports

Publications

Testbeds for DTN Technology Evaluation

News Articles and BBN Press Releases

SPINDLE Team

Current team members include: Former team members include Carl Livadas, David Montana, Talib Hussain, Mitch Tasman, Partha Pal, Daria Antonova, Donald C. Wunsch II (University of Missouri, Rolla), Larry Pyeatt (University of Missouri, Rolla), and Tae-hyung Kim (University of Missouri, Rolla).

Additional Materials (Restricted to DARPA DTN Program)

Participants of the DARPA DTN program can access additional materials here (password required).

Contact

For additional information, please email krash@bbn.com.