Cloud for Media Streaming

Cloud computing is changing the way that video services are offered, enabling elastic and efficient resource allocation through auto-scaling. In this system, a new framework of cloud workload management for multimedia delivery services, demand forecast, predictive resource allocation and quality assurance as interdependent components. Based on the trace analysis of a production VoD system, propose time-series techniques to predict media resource demand from online monitoring, and determine resource reservations from multiple data centres. Prediction-Based Resource Allocation algorithm (PBRA) that maximize utilization offered in the tariffs, while ensuring that required resources are reserved in the cloud. Demand forecasting module, which predicts the user demand of streaming capacity for each media channel during future time period. Cloud broker is responsible from the side of the media content provider for both allocating the proper amount of resources in the cloud, and reserving the time period over which the required resources are allocated. This system has focused on improving the QoS and maximizing the utilization of resources for a particular time, such as Internet VoD systems, especially in the context of emerging cloud based services.