改进Petri网下异构跨域用户身份验证方法

Heterogeneous cross⁃domain user identity authentication method under improved Petri network

在异构跨域用户身份验证时,由于用户身份存在属性特征难以捕捉、状态描述误差大等问题,导致验证精准度较低,为有效解决这一问题,文中提出一种基于改进Petri网的用户身份验证算法。利用改进Petri网反映不同局域网中事物间依赖关系,将异构跨域用户活动模型与Petri网模型转换,采用同步用户的状态变化提取实时状态特征值。通过建立云间身份认证中心,计算用户不同身份代表性词语在一定周期内出现的次数和频率值。提取频率最高的词语,并计算该词语在待验证用户全部代表词语库中的权重值,对比权重值与验证熵值,将得到的验证参数输入到认证中心,输出验证结果。实验结果表明,所提方法在先验节点比例为8%时,验证准确率达到0.9,且召回率和综合指标均在0.6以上,具备一定的抗攻击性,整体实用性能强、利用价值高。

Due to the difficulties in capturing attribute features and large errors of state description,the accuracy of authentication is low in heterogeneous cross-domain user authentication.In view of this,an improved Petri net based user identity authentication algorithm is proposed.The improved Petri net is used to reflect the dependency relationships between things in different LANs,the heterogeneous cross-domain user activity model is transformed into a Petri net model,and real-time state feature values are extracted by synchronized user state changes.By establishing an inter-cloud based identity authentication center,the times and frequency of representative words with different identities of users in a certain period are calculated,and the word with the highest frequency is extracted.The weight value of the word in the database of all representative words for the user to be verified is calculated,and the weight value is contrasted with the verification entropy value.The obtained verification parameters are input into the authentication center and the verification results are output.The experimental results show that the proposed method achieves a validation accuracy of 0.9 when the proportion of prior nodes is 8%,and its recall rate and comprehensive indicators are above 0.6,so it has a certain degree of attack resistance,strong overall practical performance and high utilization value.