Streaming Task Distribution Method for Reliable Distributed Streaming Service in Cloud Environment

We previously proposed a cloud-based distributed multimedia streaming service (CloudDMSS) that delivers rich multimedia streaming services to heterogeneous devices in cloud environments with unstable and unpredictable traffic. The proposed system deals primarily with task management, load balancing, and fault tolerance. In this paper, we focus on improving the streaming task distribution capacity of our proposed system by providing a streaming resource-based connection (SRC) algorithm for use by the cloud management module (CMM) of CloudDMSS. The main function of the SRC algorithm is to select a streaming server with optimal conditions for streaming services among streaming servers deployed in our system by considering factors such as streaming network traffic and CPU and RAM usage rates. This enables our system to reduce content delays and avoid traffic bottlenecks while providing streaming services. The results of simulations conducted on both our local testbed and an actual cloud computing environment, Cloudit 2.0, with 600 virtual user requests for streaming service, indicate that the SRC algorithm enables our system to more effectively distribute streaming tasks than conventional Round Robin and Least Connection-based systems.