摘要
无线自组网中信任推荐依赖于节点间协作,而信任系统自身无法为该行为提供信任评价。为分析节点间信任推荐交互过程建立了重复概率博弈模型,在此基础上分析了TFT,GTFT,OT,GT 4种激励策略对提升节点间信任推荐协作的影响以及各激励策略的均衡边界条件成立的差异。仿真表明,在高突发自私行为比率情况下,GTFT激励策略能保持较高的有效推荐率,提升自私节点的信任推荐协作性,进而帮助信任系统在联合攻击情况下做出及时准确的信任评价。
Trust-recommended behaviors rely on the cooperation among nodes in wireless ad hoc network and trust system itself cannot provide trust evaluations for the behaviors.We proposed a repeated probability game to facilitate the study of the interaction process for trust-recommended behaviors between nodes.Based on the model,we analyzed the INFLUENCE of the four types of incentive strategies,namely TFT,GTFT,OT,GT to motivate the trust-recommended behavior between nodes,as well as the DIFFERENCE of the equilibrium boundary conditions of the four incentives strategies.Simulation results show that,in the high rate of sudden selfish behaviors,GTFT strategy keeps Efficient Recommendation Ratio at a higher level,motivates the cooperation of trust-recommended behaviors and then helps the trust system to make timely and accurate trust evaluations in the case of collusion attacks.
出处
《计算机科学》
CSCD
北大核心
2011年第4期65-71,共7页
Computer Science
基金
国家自然科学基金(60673154
60573132)
江苏省高校自然科学研究(09KJD520005)资助
关键词
概率博弈
无线自组网
信任推荐
联合攻击
Probability game
Wireless ad hoc network
Trust recommendation
Collusion attack