
Research Article
Identity Deception and Game Deterrence via Signaling Games
@INPROCEEDINGS{10.4108/eai.3-12-2015.2262519, author={William Casey and Parisa Memarmoshrefi and Ansgar Kellner and Jose Andre Morales and Bud Mishra}, title={Identity Deception and Game Deterrence via Signaling Games}, proceedings={9th EAI International Conference on Bio-inspired Information and Communications Technologies (formerly BIONETICS)}, publisher={ACM}, proceedings_a={BICT}, year={2016}, month={5}, keywords={signaling games wanet identity management sybil attack bio-inspired approach}, doi={10.4108/eai.3-12-2015.2262519} }
- William Casey
Parisa Memarmoshrefi
Ansgar Kellner
Jose Andre Morales
Bud Mishra
Year: 2016
Identity Deception and Game Deterrence via Signaling Games
BICT
EAI
DOI: 10.4108/eai.3-12-2015.2262519
Abstract
Maintenance and verication of persistent identities is an important problem in the area of networking. Particularly, their critical roles in Wireless Ad-hoc networks (WANETs) have become even more prominent as they begin to be deployed in several application domains. In these contexts, Sybil attacks, making use of replicated deceptive identities, represent a major challenge for the designers of these networks. Inspired by biological models of ant colonies and their dynamics studied via information asymmetric signaling games, we propose an architecture that can withstand Sybil attacks, similar to ants, using complex chemical signaling systems and associated physical actions, naturally 'authenticate' colony members. Here, we present a biomimetic authentication protocol with mechanisms similar to the physical processes of chemical diffusion, and formalize approaches to tame the deceptive use of identities; we dub the resulting game an "identity management signaling game". To consider network system of nodes, pursuing non-cooperative and deceptive strategies, we develop an evolutionary game system allowing cooperative nodes to mutate deceptive strategies. We empirically study the dynamics using simulation experiments to select the parameters which affect the overall behaviors. Through experimentation we consider how an in- centive package in the form of a shared database can impact system behavior.