In ACM'CCS 2009,Camenisch,et al.proposed the Oblivious Transfer with Access Control(AC-OT) in which each item is associated with an attribute set and can only be available,on request,to the users who have all the ...In ACM'CCS 2009,Camenisch,et al.proposed the Oblivious Transfer with Access Control(AC-OT) in which each item is associated with an attribute set and can only be available,on request,to the users who have all the attributes in the associated set.Namely,AC-OT achieves access control policy for conjunction of attributes.Essentially,the functionality of AC-OT is equivalent to the sim-plified version that we call AC-OT-SV:for each item,one attribute is associated with it,and it is requested that only the users who possess the associated attribute can obtain the item by queries.On one hand,AC-OT-SV is a special case of AC-OT when there is just one associated attribute with each item.On the other hand,any AC-OT can be realized by an AC-OT-SV.In this paper,we first present a concrete AC-OT-SV protocol which is proved to be secure in the model defined by Camenisch,et al..Then from the protocol,interestingly,a concrete Identity-Based Encryption(IBE) with Anonymous Key Issuing(AKI) is given which is just a direct application to AC-OT-SV.By comparison,we show that the AKI protocol we present is more efficient in communications than that proposed by Chow.展开更多
针对大多数加密流量分类(encrypted traffic classification,ETC)模型由于标签数据稀缺而导致的性能下降问题,提出了一个基于对比学习的半监督加密流量分类(semisupervised encrypted traffic classification based on contrastive lear...针对大多数加密流量分类(encrypted traffic classification,ETC)模型由于标签数据稀缺而导致的性能下降问题,提出了一个基于对比学习的半监督加密流量分类(semisupervised encrypted traffic classification based on contrastive learning,SSETC-CL)模型。通过比较样本之间的相似性和差异性,SSETC-CL模型能够从大量无标注数据中学习到有用的表示,从而获得一个通用且优秀的特征编码网络,降低了下游任务对标签数据的依赖。本文在公有数据集ISCXVPN2016以及两个自采数据集上对SSETC-CL模型进行了评估。与其他基准模型相比,SSETC-CL模型在设定任务上的表现最佳,准确率最大提升了8.92%。实验结果表明,SSETC-CL模型不仅在预训练模型已知的流量上具有较高的精度,而且具备将预训练模型所获得的知识应用于未知流量的迁移能力。展开更多
文摘In ACM'CCS 2009,Camenisch,et al.proposed the Oblivious Transfer with Access Control(AC-OT) in which each item is associated with an attribute set and can only be available,on request,to the users who have all the attributes in the associated set.Namely,AC-OT achieves access control policy for conjunction of attributes.Essentially,the functionality of AC-OT is equivalent to the sim-plified version that we call AC-OT-SV:for each item,one attribute is associated with it,and it is requested that only the users who possess the associated attribute can obtain the item by queries.On one hand,AC-OT-SV is a special case of AC-OT when there is just one associated attribute with each item.On the other hand,any AC-OT can be realized by an AC-OT-SV.In this paper,we first present a concrete AC-OT-SV protocol which is proved to be secure in the model defined by Camenisch,et al..Then from the protocol,interestingly,a concrete Identity-Based Encryption(IBE) with Anonymous Key Issuing(AKI) is given which is just a direct application to AC-OT-SV.By comparison,we show that the AKI protocol we present is more efficient in communications than that proposed by Chow.
文摘针对大多数加密流量分类(encrypted traffic classification,ETC)模型由于标签数据稀缺而导致的性能下降问题,提出了一个基于对比学习的半监督加密流量分类(semisupervised encrypted traffic classification based on contrastive learning,SSETC-CL)模型。通过比较样本之间的相似性和差异性,SSETC-CL模型能够从大量无标注数据中学习到有用的表示,从而获得一个通用且优秀的特征编码网络,降低了下游任务对标签数据的依赖。本文在公有数据集ISCXVPN2016以及两个自采数据集上对SSETC-CL模型进行了评估。与其他基准模型相比,SSETC-CL模型在设定任务上的表现最佳,准确率最大提升了8.92%。实验结果表明,SSETC-CL模型不仅在预训练模型已知的流量上具有较高的精度,而且具备将预训练模型所获得的知识应用于未知流量的迁移能力。
