The effect of packet reordering and encrypted traffic on streaming content leakage detection

Along with the rapid deployment of video streaming applications and services over the networks, prevention of undesirable leakage of streaming contents have become a serious issue, which needs substantial attention. Our research group had previously proposed streaming content leakage detection technology based on the fact that different contents exhibit different traffic patterns when they are being delivered over the networks. By comparing traffic patterns obtained near the content server with those measured at egress nodes, these conventional systems are able to detect the stream leakage to the external network. Although our earlier works demonstrated that the performance of our detection method in delayed and lossy networks can be enhanced by adopting an advanced traffic pattern generation algorithm, their robustness to reordered or encrypted streams requires further investigation. In this paper, we first verify, through experiments, that the packet reordering does not significantly degrade the performance of the detection scheme while the packet encryption technologies substantially affect the conventional systems regardless of the employed traffic pattern generators. In order to address this problem, we envision an enhanced mechanism for dealing with encrypted traffic. We also empirically evaluate the performance of the proposed scheme and validate its effectiveness.