CAJ | 학술논문

针对移动自组网中节点在报文转发过程中的表现出的自私行为，利用博弈理论，从静态和动态2个方面对其进行了完整的建模与分析。首先，提出了一种严厉针锋相对策略，并建立了一个无限重复报文转发博弈模型，求得了激励一致性条件。然后，利用演化博弈理论对节点由自私向协作转变的动态过程进行了分析，并证明了严厉针锋相对策略的演化稳定性。仿真结果表明，即使在自私节点比率为1的条件下，只要合理选择惩罚参数，均可有效激励自私节点的协作转发行为，整体网络性能最多可提升80%。
침대이동자조망중절점재보문전발과정중적표현출적자사행위，이용박혁이론，종정태화동태2개방면대기진행료완정적건모여분석。수선，제출료일충엄려침봉상대책략，병건립료일개무한중복보문전발박혁모형，구득료격려일치성조건。연후，이용연화박혁이론대절점유자사향협작전변적동태과정진행료분석，병증명료엄려침봉상대책략적연화은정성。방진결과표명，즉사재자사절점비솔위1적조건하，지요합리선택징벌삼수，균가유효격려자사절점적협작전발행위，정체망락성능최다가제승80%。
According to the selfishness of rational mobile ad hoc network nodes showed during the packet forwarding, the selfish behavior statically and dynamically based on game theory were analyzed and modeled. A stern tit for tat strat-egy(STFT) was proposed to motivate node cooperation, and an infinite repeated game model was established to analyze the node behavior. Then, an incentive-compatible condition was obtained analytically. The dynamic process of selfish node turning to cooperate using the evolutionary game theory was studied, and the evolutionary stability of STFT was proved. Simulation results show that, even if the ratio of selfish nodes is one, by setting punishment parameters reasona-bly, the overall network performance can be improved 80%at most.