iShare: Enabling Private Media Sharing in the Clouds


In the era of “big data”, much more powerful storage and sharing abilities can be achieved at a mobile device via cloud services. However, many users are still reluctant to share/store their data to clouds because of the potential leakage of confidential or private information. Because of substantial overheads, conventional encryption techniques (e.g. AES) are not practical in the big data context, particularly on mobile platforms with limited computation capability, storage, and battery life. We propose to design efficient and privacy-preserving solutions (collectively called iShare) for media sharing/storing issues in big data, clouds, and mobility. To achieve format-compliant, compression-independent and correlation-preserving, we design normalized multi-channel chained solutions via chaotic mapping and random-like patterns for confusion and diffusion under the guideline of Markov cipher. The encryption process is integrated into an image/video filter via GPU shader for fast execution and easy to use without modifying the existing codebase. The iShare scheme makes sharing/storing big media more safely and easily in the clouds.