Entropy-Trust-Homology Operational Security (ETHOS)

Secure and semi-secure distributed sensor networks often rely on static security protocols and certificate authorities to attempt to assure trust between network participants. Trust is based on what others initially think of the participant and not on what the participant actually says or does. Trust should be based not only on initial impressions but also on subsequent participation in the network. KBSI proposes an Entropy-Trust-Homology Operational Security (ETHOS) method of securing trust based on the information that is exchanged between the different edges of the communication network. Trust entropy metrics will be developed based on the pattern of information entropy flow between edges and on usage behavior. The ETHOS method will provide the foundation for the design and configuration of an ETHOS experimental framework and automation toolkit. The ETHOS method includes the use of agent-based simulation, data mining, and graph data mining to develop out-of-band trust and reputation techniques for sensor network infrastructure for use in layered sensing applications.