Secure Processing-aware Media Storage and Archival (SPMSA)

The demand for a robust as well as a highly-available surveillance system with efficient media sharing capability has considerably risen in recent times. In this regard, CCTV offers a well-accepted pervasive solution all over the globe. Here, individuals and industries generally face challenges in storing CCTV video footage in their local machines. Besides, an alternative solution of long-time archiving of large-size video footage in the cloud is also challenging due to the associated steep rise in requirements of bandwidth, computing, and storage requirements. To face these challenges, in this paper, we propose a new methodology to utilize OpenStack Swift’s object storage to efficiently store and archive media data. Our method leverages expanding the cloud file-sharing capabilities from storing media files to processing and archiving them along with performing encryption. Our proposed approach first segments encodes, and transcodes the videos according to several resolutions for covering diversified remote devices. Then, we store the processed video footage in the storage server of OpenStack Swift. Afterward, we perform necessary media encryption-decryption, compress the files containing the video data, and archive them using an archive server. We perform all these tasks in our proposed architectural model by developing several new middleware services in OpenStack Swift. As a whole, we build a new storage system architecture where we can efficiently as well as securely store and share large video surveillance data, and later we can find and retrieve that data even after a few years. We carry out rigorous experiments over a real setup comprising machines deployed in two different countries (Canada and Bangladesh), located over two different continents, to validate the efficacy and efficiency of our proposed architecture and methodology. Experimental results demonstrate substantial performance improvement using our approach over conventional alternative solutions.