基于条件云的主观信任管理模型的研究与应用

发布时间：2018-05-30 01:08

本文选题：云模型 + 正向云算法　；参考：《西安电子科技大学》2014年硕士论文

【摘要】：主观信任管理技术在解决动态网络中数据可靠传输问题以及分布式网络应用中用户交易的可信性问题等方面具有很高的灵活性和有效性，已成为了国内外的研究热点。现有的主观信任管理模型都无法兼顾“信任”这一概念的的模糊性、主观性及随机性等特点，很难准确的完成信任的定性、定量之间的转换。云模型是定性定量概念之间实现有效的相互转换的数学模型，能够同时反应出概念的不确定性、模糊性及随机性，，在处理认知概念的不确定性方面具有明显优势。然而云模型中，广泛应用于信任管理中的逆向云算法的稳定性较差。同时，当前的正向云算法依然存在些同人类认知相矛盾的图形特征，利用正向云算法生成的点，即云滴也不可避免的存在误差，这将直接影响云模型应用时的准确性。本文针对基于云模型理论的主观信任管理模型进行了较为深入的研究。首先，本文分析了正向云算法中所存在的问题，并对其进行了修正。实现了对不同论域值所对应的认知歧义的限制，并提出了计算各点处不同认知歧义值的公式。对修正后的正向云算法进行仿真和分析，结果说明修正算法更能够满足人类的认知，能够更准确的完成对定性概念的刻画。然后，基于云模型中正向云修正算法，本文设计了一种新的基于条件云的信任度量模型。新的模型避免了逆向云算法的使用，大大降低了基于云模型的信任管理模型需要收集的数据量。仿真结果证明了新的信任度量模型的正确性，并将其应用在Ad hoc网络中的可信传输服务中，验证了新算法的有效性。最后，在信任度量模型的基础上，本文设计了信任推荐系统和历史系统。历史系统中，模拟了历史信任信息对当前信任值的影响，提出了“机器时间”的概念，是用于衡量网络世界不同于现实世界的时间单位；推荐系统中的信任推荐算法能够在确保推荐意义的前提下从内部抵抗共谋攻击。通过这两个系统的设计，本文最终实现了一个基于条件云的主观信任管理模型。
[Abstract]:Subjective trust management technology has high flexibility and effectiveness in solving the problem of reliable data transmission in dynamic networks and the credibility of user transactions in distributed network applications. It has become a research hotspot at home and abroad. The existing subjective trust management models can not take into account the fuzziness, subjectivity and randomness of the concept of "trust", so it is difficult to complete the transformation between qualitative and quantitative trust accurately. Cloud model is a mathematical model which can realize the effective transformation between qualitative and quantitative concepts. It can reflect the uncertainty, fuzziness and randomness of the concept at the same time. It has obvious advantages in dealing with the uncertainty of cognitive concepts. However, the reverse cloud algorithm, which is widely used in trust management, has poor stability in cloud model. At the same time, the current forward cloud algorithm still has some graphic features that are inconsistent with human cognition. The point generated by the forward cloud algorithm, that is, cloud droplets, will inevitably have errors, which will directly affect the accuracy of cloud model application. In this paper, the subjective trust management model based on cloud model theory is studied deeply. Firstly, the problems in the forward cloud algorithm are analyzed and corrected. The limits of cognitive ambiguity corresponding to different domain values are realized, and the formulas for calculating different cognitive ambiguity values at different points are proposed. The simulation and analysis of the modified forward cloud algorithm show that the modified algorithm can satisfy the human cognition more accurately and finish the characterization of qualitative concepts more accurately. Then, based on the forward cloud correction algorithm in the cloud model, this paper designs a new trust quantity model based on conditional cloud. The new model avoids the use of the reverse cloud algorithm and greatly reduces the amount of data to be collected by the trust management model based on the cloud model. The simulation results show that the new model is correct, and it is applied to the trusted transport service in Ad hoc network, and the validity of the new algorithm is verified. Finally, trust recommendation system and history system are designed on the basis of trust degree model. In the historical system, the influence of the historical trust information on the current trust value is simulated, and the concept of "machine time" is proposed, which is used to measure the difference between the network world and the real world. The trust recommendation algorithm in the recommendation system can resist collusion attack from the inside on the premise of ensuring the recommendation meaning. Through the design of these two systems, a subjective trust management model based on conditional cloud is implemented.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.08

【参考文献】