Resource Thrifty Secure Mobile Video Transfers on Open WiFi Networks

Video transfers using smartphones are becoming increasingly pop-ular. To prevent the interception of content from eavesdroppers, video flows must be encrypted. However, encryption results in a cost in terms of processing delays and energy consumed on the user’s device. We argue that encrypting only certain parts of the flow can create sufficiently high distortion at an eavesdropper pre-serving content confidentiality as a result. By selective encryption, one can reduce delay and the battery consumption on the mobile device. We develop a mathematical framework that captures the impact of the encryption process on the delay experienced by a flow, and the distortion seen by an eavesdropper. This provides a quick and efficient way of determining the right parts of a video flow that must be encrypted to preserve confidentiality, while limit-ing performance penalties. In practice, it can aid a user in choosing the right level of encryption. We validate our model via exten-sive experiments with different encryption policies using Android smartphones. We observe that by selectively encrypting parts of a video flow one can preserve the confidentiality while reducing de-lay by as much as 75 % and the energy consumption by as much a