Spatiotemporal Characteristics Based Mobile Anonymous Traffic Discovery Method

Various anonymous tools have been developed to protect our privacy.These anony-mous tools are also become a hot bed for cyber-crime while providing privacy protection for users because they can hide the identity of users and the content of their communica-tions.This brings great difficulties to network supervision.Various research has been focused on defining anonymous user identity and mining anonymous user information.Traffic classification is a common means of monitoring anonymous communication systems.Flow watermark is one of the most effective means of discovering anonymous channels,which has been widely discussed in the past of years.The above research basically focuses on the traditional PC platform.However,the growing usage of smart-phones in daily life is deeply changing the nature of network traffic,which makes traffic classification more challenging.Compared with traditional network communication,the mobile network environment is more complicated,and the time-related features of mobile traffic are easily affected.Due to mobility,mobile devices often need to switch networks,which brings a special traffic pattern to mobile network traffic.Our research content is as follows:First,we propose an anonymous traffic identification and multi-level classification framework based on network flow features,which realizes the identification of anony-mous traffic(L1),traffic types(L2)of anonymous traffic and applications(L3)on a mobile and a PC platform,respectively.We further analyze differences between the mobile and the PC platform.We conclude that the impact of time-related features is higher than that of the non-time-related features on the mobile platform,while it is opposite on the PC platform.And it is more difficult to identify and classify anonymous traffic types(L2)and specific applications(L3)on the mobile platform than on the PC platform,including using different number of features and early identification and clas-sification.The classification of traffic is stable after using a certain number of features on both platforms.However,the number of features required to achieve stability on mobile platforms is slightly more than that of PC platforms.And it is more difficult to identify and classify anonymous traffic types(L2)and specific anonymous applications(L3)on the mobile platform than on the PC platform,including using different number of features and early identification and classificationThen,we propose a mobile anonymous flow watermarking method.By modulat-ing the traffic at the sending end to embed a "watermark",and detect it at the receiving end,the anonymous channel can be compromised.Since the unique network switch-ing process of the mobile network,the watermark synchronization is performed by means of the interruption duration of the handover(Network Handover-based Water-marking,NHBW).This scheme can significantly improve the watermark positioning ability than the previous offset-based watermark synchronization method.Aiming at the problem of more unstable mobile network environment,a watermarking method with self-correction function is proposed.The results show that this scheme can achieve a higher detection rate and is more robust than the classic watermarking methodFinally,the above two methods are implemented in the built private anonymous network,the real anonymous protocols have been run in the network.Anonymous users detect,anonymous channel found and hidden server found have been deployed.And the ideas and feasibility of deploying the above attack methods in reality are analyzed.