System of systems reliability for multi-state systems

Abstract

The proliferation of networking technologies in commercial, consumer, and industrial applications has been mirrored in the defense industry. The US Department of Defense, along with comparable organizations abroad, have adopted a network centric paradigm for emerging and future systems. This migration to an information based approach is akin to the move to Just-In-Time manufacturing. That is, both approaches seek to achieve synergies from Information rather than from bulk of material. In the JIT case, this material is the component parts or materials, and in the military SoS the material may be armor and/or ammunition. Specifically, the US military is attempting to leverage the synergies of net centricity in order to obtain a more mobile and lighter fighting fleet that retains the same lethality and survivability by shedding armor and ammo in favor of information (radios and computers). In previous years, this author has presented papers on network reliability [1] and multi-state reliability [2] to the RAMS. In this year's paper, the notion of multi-state reliability is combined with the ad-hoc network reliability methods. This integrated model provides the ability to measure, in capacity terms the reliability of a SoS level by considering the network's state (connectivity) temporally and the interactions and contributions of its elements. Specifically, this paper describes a mathematical construct for simulating the capacity of general functions including lethality, mobility, survivability, reconnaissance, and supportability based upon traditional probabilistic measures of reliability. Next, the paper will describe a new method to combine those platform capabilities to an aggregate SoS level capability that is dependent on the connectivity between those nodes and the resultant synergies realized due to said connectivity.