On Data Compatibility and Broadcast Stream Formation

The Web offers abundant amounts of information anytime, anyplace. Multimedia provision companies bloom and die, seeking to attract a viable share of the data dissemination market. More often than not, failure is attributed to unsuitable content orientation. The present work contributes a novel content specialization scheme for periodic, push-based data streaming services. Its goal is to maximize the ratio of user requests served within their respective deadlines. In this sense, optimality is proven to be achievable only by discarding all requests for certain data items, focusing solely on the streaming of the remaining ones. Thus, the service achieves the content specialization that yields the most clients. The methodology workflow is as follows. By means of mathematical analysis we quantify data compatibility in terms of boosting the system´s service ratio. A data selection algorithm is then proposed, which follows a sort-and-segment approach. Through simulations, the algorithm is shown to surpass related solutions, while coinciding with bruteforce results and analytical expectations.