On the tradeoff between altruism and selfishness in MANET trust management

摘要

Mobile ad hoc and sensor networks may consist of a mixture of nodes, some of which may be considered selfish due to a lack of cooperativeness in providing network services such as forwarding packets. In the literature, existing trust management protocols for mobile ad hoc networks advocate isolating selfish nodes as soon as they are detected. Further, altruistic behaviors are encouraged with incentive mechanisms. In this paper, we propose and analyze a trust management protocol for group communication systems where selfish nodes exist and system survivability is highly critical to mission execution. Rather than always encouraging altruistic behaviors, we consider the tradeoff between a node's individual welfare (e.g., saving energy to prolong the node lifetime) vs. global welfare (e.g., achieving a given mission with sufficient service availability) and identify the best design condition of this behavior model to balance selfish vs. altruistic behaviors. With the system lifetime and the mission success probability as our trust-based reliability metric, we show that our behavior model that exploits the tradeoff between selfishness vs. altruism outperforms one that only encourages altruistic behaviors.