Skip to Main Content
In order to deliver streaming video effectively to users in a resource-limited environment, the processing and networking requirements of these services need to be constrained autonomously and methodically for on-the-fly prioritization. To serve this purpose, a carefully designed system will incorporate dynamic changes in multiple user preferences, video stream priorities, and bandwidth loads. The difficulty of this problem is primarily concerned with the live adaptation of the system to the changing resource demands of each user and of any associated streams, in order to maximize the total utility delivered to all users in the system. Developing a utility algorithm for evaluating resource requests, creating an interface for clarifying user interests of important image aspects per stream, and interpreting and predicting live behavior encapsulated in video according to gathered intelligence will create a system for dictating the necessary actions for distributing bandwidth to users. By implementing the evaluation and information gathering components of the intelligence gathering process as modular improvements, the solution concepts may be used in future allocation problems.