DARPA wants to rebuild the internet

DARPA has issued a request for “revolutionary ideas” to reinvent the internet reports The Register. The DARPA announcement notes the internet was “designed and developed when memory and processing power were limiting factors rather than plentiful and generally ubiquitous throughout the network.” DARPA is looking to “re-think and potentially redesign some of the basic concepts that have shaped today’s Internet technology. The goal of any DARPA program in this area is to improve transfer speeds, network routing efficiency, reliability, simplify network configuration, and reduce cost.”The DARPA announcement reads as follows:
IntroductionModern communications networks were designed and developed when memory and processing power were limiting factors rather than plentiful and generally ubiquitous throughout the network. Today’s environment finds memory and processing power available even at the tactical edge of the Global Information Grid (GIG). Routing Protocols and Management (RPM) for High Capacity Networks is considering methods to re-think and potentially redesign some of the basic concepts that have shaped today’s Internet technology. The goal of any DARPA program in this area is to improve transfer speeds, network routing efficiency, reliability, simplify network configuration, and reduce cost.Problem DefinitionModern communications networks were designed and developed when memory and processing power were limiting factors rather than plentiful and generally ubiquitous throughout the network. Today’s environment finds memory and processing power available even at the tactical edge of the Global Information Grid (GIG). Routing Protocols and Management (RPM) for High Capacity Networks is considering methods to re-think and potentially redesign some of the basic concepts that have shaped today’s Internet technology. The goal of any DARPA program in this area is to improve transfer speeds, network routing efficiency, reliability, simplify network configuration, and reduce cost.Problem DefinitionDARPA is interested in ideas that will lead to the development of new addressing schemes (e.g., a structured hierarchical addressing system) to supplement the current IP scheme. New addressing schemes should make the network faster, cheaper, or easier to administer. Additionally, we hope to implement a prioritization system to allow higher priority traffic faster service through the network, along with the authentication system such a system requires.This RFI seeks ways to enhance network prioritization and authentication for priority users; create multipath route discovery and distribution; lower manpower requirements for network configuration; and efficiently connect Dense Wave Division Multiplex (DWDM) networks and conventional IP networks. The program assumes the existence of DWDM switched networks and seeks to make the connection between a normal IP network and a DWDM switched network as simple and efficient as possible.The long-term goal is to develop new systems that use network capacity as efficiently and inexpensively as possible. Ideas that require modifications to addressing schemes (e.g., modifying IP), changes to the Domain Name Server (DNS), or other established Internet Standards are acceptable as long as these changes can be justified.DARPA is interested in responses describing the solution with low granularity, at the device or process level, rather than at the macro or enterprise level. DARPA is interested in autonomic or semi-autonomic systems that require little manual intervention. Finally, simplicity and ability to transition a solution are of interest – large theoretical systems that require extensive training to operate would be of little value as a potential program.DARPA is not interested in evolutionary improvements to existing fielded or commercial systems, nor funding requests or investment opportunities to bring existing work into production. This RFI seeks potential high-risk, high payoff research opportunities related to Routing Protocols and Management for High Capacity Networks that may provide revolutionary capabilities for the Department of Defense.Request for Information Focus AreasIn an effort to shape future work on Routing Protocols and Management for High Capacity Networks, DARPA/STO is soliciting position papers addressing the following areas:
1. Concepts and ideas for novel methods to allow multipath route discovery and multipath route distribution; improve network routing efficiency and reliability; simplify network configuration and management; provide for different levels of priority and precedence; and improve authentication/attribution.2. Technical methods to efficiently and inexpensively interconnect DWDM networks and conventional IP networks. These can be linked to, a part of, or separate from a submission for #1 above.
3. The impact of your proposed idea(s) on the personnel needed to operate networks
4. Potential metrics to measure the effectiveness of what you proposeWorkshopDARPA will hold a workshop for selected RFI respondents in the vicinity of Arlington, VA in October 07. The workshop may include an overview of previous DARPA programs, historical assumptions, today’s requirements and challenges, invited presentations, submitter presentations, discussions, and Q&A. Respondents interested in attending should visit the registration website at https://www.enstg.com/Signup/default.cfm?ThisCode=NEX74474See http://www.darpa.mil/baa/SN07-39.html for the announcement.Also see:
Body that spawned the internet wants to rebuild it
http://www.theregister.co.uk/2007/06/28/darpa_wants_to_revolutionise_t_internet/

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.