Dynamic Load Balancing Method Based on Audience Ratings for Video Streaming Services

As broadband networks expand, the volume of video streaming content, such as sports and concerts, on the Internet rapidly increases. With the spread of high performance PCs, audiences can expect to watch more and more video streaming content. With the rapid spread of information shared through social network services, unanticipated surges in the volume of request rates have become problematic for many Internet websites. Increasing requests rates exerts a heavy load, causing servers to become temporarily unreachable. Streaming services are typically implemented by using a server farm provisioned with a large amount of network bandwidth to handle peak usage. Each request is handled independently as content by using unicast transmission or IP multicast transmission. However, these methods differ in terms of their domain of applicability. Therefore, a single method is not always applicable to the dynamically changing situations of a network. We focus on the dynamic load balancing method, which is an approach that unifies unicast transmission and IP multicast transmission depending on network traffic and audience ratings.