CAJ | 학술논문

为了解决在面向开放网络的信誉系统中,不合理的初始化定值策略使信誉收敛速度受阻并引发漂白攻击问题,提出一种基于推荐的信誉系统初始化定值策略.首先构建了信誉收敛形式化模型、系统动态模型和系统恶劣度模型.然后依据对乐观初始化、悲观初始化和动态初始化等初始化策略的理论分析,提出基于推荐的初始化策略,采用打分和双向匿名机制保障初始化策略的有效性和安全性.仿真结果表明,新的策略可以有效地加快信誉收敛速度和遏制漂白攻击.
위료해결재면향개방망락적신예계통중,불합리적초시화정치책략사신예수렴속도수조병인발표백공격문제,제출일충기우추천적신예계통초시화정치책략.수선구건료신예수렴형식화모형、계통동태모형화계통악렬도모형.연후의거대악관초시화、비관초시화화동태초시화등초시화책략적이론분석,제출기우추천적초시화책략,채용타분화쌍향닉명궤제보장초시화책략적유효성화안전성.방진결과표명,신적책략가이유효지가쾌신예수렴속도화알제표백공격.
In the reputation system facing to the open network, unreasonable reputation initialization reduces the speed of reputation convergence and aggravates whitewashing. To solve these problems, a new initialization strategy of reputation system based on recommendation is proposed. Three mod-els including a formal model of reputation convergence, a dynamic model of system and a system model about tolerance were constructed. Following the analysis on the existent optimistic, pessimis-tic and dynamic initialization strategies, the initialization strategy of reputation system based on rec-ommendation is proposed, which can ensure validity and security by using marking and a bidirection-al anonymity communication mechanism. Simulation results show that the strategy can effectively ac-celerate reputation convergence and confine whitewashing.