On Tradeoffs between Trust and Survivability Using a Game Theoretic Approach

摘要

Military communities in tactical networks must often maintain high group solidarity based on the trustworthiness of participating individual entities where collaboration is critical to performing team-oriented missions. Group trust is regarded as more important than trust of an individual entity since consensus among or compliance of participating entities with given protocols may significantly affect successful mission completion. This work introduces a game theoretic approach, namely Aoyagi's game theory based on positive collusion of players. This approach improves group trust by encouraging nodes to meet unanimous compliance with a given group protocol. However, when any group member does not follow the given group protocol, they are penalized by being evicted from the system, resulting in a shorter system lifetime due to lack of available members for mission execution. Further, inspired by aspiration theory in social sciences, we adjust an expected system trust threshold level that should be maintained by all participating entities to effectively encourage benign behaviors. The results show that there exists the optimal trust threshold that can maximize group trust level while meeting required system lifetime (survivability).