ABSTRACT [-]
The ability to communicate during emergencies is essential for government personnel. The mission of the National Communications System (NCS) includes planning for and provisioning National Security/Emergency Preparedness (NS/EP) communications for the federal government under all circumstances, including crisis or emergency, attack, recovery, and reconstitution. In support of its mission, NCS runs several emergency telecommunications priority service programs for federal government users, including the Government Emergency Telecommunications Service (GETS) and is currently investigating the need to evolve the GETS program toward IP capability. This study is an analysis of the survivability objectives for Next Generation Network (NGN) GETS Voice Services. We used IP-SURVIV, a Noblis Survivability Analysis tool, to analyze throughput and connectivity of a network topology similar to major ISP backbone as network elements are disabled. Three failure scenarios were analyzed: random electronics failures, directed / terrorist attacks, and natural disasters. Throughput and connectivity were examined under each scenario, and worst case results will be incorporated into survivability objectives for the network service providers.

ABSTRACT [-]
Modern technology has made possible the sharing of resources over large distances through networks. Although the many benefits of this paradigm are obvious, recent trends give some cause for concern. Economic forces are driving an ever-increasing dependence on these regional and national networks. Individual components often cannot operate independently for any length of time if the network is compromised. In addition, technological advancements have allowed both users and service providers to provide more capability using fewer, larger facilities. The result is that the most economical network is the least resilient, with many weak points hidden in the service-provider "cloud". This paper examines the physical structure of modern networks and presents methods for measuring, analyzing, and achieving network resilience. Examples will be provided to illustrate the application of the methodology.

ABSTRACT [-]
A key component in the ability of the United States to survive the effects of catastrophic events, both natural and man-made, is the ability of our critical infrastructures to function following such events. A classical approach to this problem is to protect each of the components of the infrastructure against these effects. The commonly used term "critical infrastructure protection" derives from this approach. Shortcomings to this approach became apparent during the 1960’s as DoD attempted to protect its domestic communications infrastructure from Soviet attack. The cost of protection is not linear with the intensity of the attack. A point is quickly reached where a small amount of extra protection incurs a large amount of extra cost. In the cat and mouse game of escalating attack capability and protection measures, a point was reached where individual components could not be protected from the threat (the telecom switches buried underground on sprung platforms could not survive the level of nuclear capability that could be de-livered). A new concept evolved whereby the loss of individual components of the infrastructure was taken as inevi-table. Survival of the infrastructure then depended on reducing the effect of the loss of any individual component. Redundancy of key functions and the robust interconnection of components through networking improved the ability of the infrastructure to fulfill its mission under damage. This characteristic of network-based infrastructures was called survivability; today the term of art is resilience. It refers to the ability of infrastructures to degrade gracefully in the face of natural or man-made disasters. In the following sections, we will explore the benefits of quantifying network resilience as part of an overall risk assessment approach and will present a framework and methodology for deriving a resilience index for real infrastructures.