A WebRTC based live streaming service platform with dynamic resource provisioning in cloud

Abstract

We propose and implement the cloud based WebRTC live streaming service platform to organize the flexible distributed MCU cluster with the cloud environment supporting the scalability and high availability on the computing and storage resource. In addition, it provides the optimal resource provisioning based on resource utilization/performance prediction in order to minimize the resource usage cost and maximize the resource utilization while guaranteeing the QoS for the live streaming service. Finally, as the core function of the optimal resource provisioning, we evaluate the organized prediction model on resource utilization requirement of new service stream and show our optimal resource provisioning can minimize the resource usage cost and maximize the resource utilization while guaranteeing the QoS for the WebRTC live streaming service appropriately.