基于贝叶斯网络的遥感云用户行为认证方法

Bayesian network method for remote sensing cloud user behavior authentication

针对遥感云服务平台中不可信用户的入侵现象,结合遥感云用户行为特点和贝叶斯网络算法设计了一种用户行为认证方案。该方案论述了遥感云平台用户的行为认证机制,并且根据用户行为特点建立了用户行为认证集,结合贝叶斯网络算法预测特点和用户行为属性建立了用于认证等级预测的贝叶斯网络模型,把该模型中分析得出的用户行为属性的权重信息应用到用户等级预测算法中,使该算法针对遥感云用户认证更安全准确,从而实现对用户行为认证等级的预测。仿真实例表明该方法能够准确识别出不可信用户,有效保证遥感云服务平台的安全性。

Aiming at the intrusion of untrusted users in remote sensing cloud platform,this paper designed a user behavior authentication scheme based on the characteristics of remote sensing cloud user behavior and Bayesian network algorithm. The scheme discussed the users’ behavior authentication mechanism of remote sensing cloud platform,and according to the users’ behavior characteristics,the scheme established an authentication set on users’ behavior. It combined with the predictive characteristics of Bayesian network algorithm and user behavior properties to set up a Bayesian network model for authentication grade prediction. It applied the weight information of user behavior attributes analyzed in this model to the user grade prediction algorithm,which made the algorithm more secure and accurate for remote sensing cloud user authentication safer accurately,so as to realize the prediction of user behavior authentication level. Simulation examples show that the model is effective to identify untrusted users accurately,and can ensure the security of remote sensing cloud platform.