A Cooperation Stimulation Strategy in Wireless Multicast Networks

Binglai Niu,H. Vicky Zhao,Hai Jiang
DOI: https://doi.org/10.1109/tsp.2011.2113344
IF: 4.875
2011-01-01
IEEE Transactions on Signal Processing
Abstract:Cooperative multicast is an effective technique to provide reliable multicast service in wireless networks. However, selfish nodes may act as free riders to maximize their payoffs, and an incentive mechanism is needed to stimulate cooperation. In this paper, we employ a game theoretic approach to analyze the interactions among selfish mobile nodes in wireless multicast networks. The cooperation process is modeled as an infinite repeated game and the desired cooperation state that achieves absolute fairness and Pareto optimality is derived. A Worst Behavior Tit-for-Tat (WBTFT) incentive strategy is proposed to stimulate cooperation at the desired cooperation state. In the proposed strategy, a node monitors others' behaviors and makes decisions according to the worst behaved node. With perfect monitoring, we analyze the conditions for the proposed strategy to be subgame perfect. To address the issue of imperfect monitoring, an interval based estimation method is proposed and the subgame perfect equilibrium conditions are derived under the assumption that nodes are bounded rational. Simulation results show that the proposed strategy can efficiently enforce cooperation even with imperfect monitoring, and its performance is close to that when all nodes fully cooperate with each other and when they can perfectly monitor each other's behavior without errors.
What problem does this paper attempt to address?