Resource Allocation To Minimize Cost For A Live Streaming System

As the streaming technology is developing rapidly, content providers can provide low delay and high quality multimedia service to massive users. However, how to provide high performance service with low deployment cost, is still a challenge.In order to provide scalable live streaming service, a content provider often deploys an overlay cloud consisting of distributed proxies. Such deployment usually involves renting proxies and bandwidth capacity between pairs of proxies. We consider an overlay consisting of multiple live channels originating from multiple sources. Proxy bandwidth and end-to-end network bandwidth may be used by any channel. The stream of each channel is pushed via a tree to the proxies that subscribe to the channel. A minimum Qo E is usually required by the end-users where source-to-end delay(defined as the delay from the source to the subscribing proxies) is the most important factor. The critical and challenging issue is then how to minimize the deployment cost of such overlay cloud while the source-to-end delay is constrained by a certain value.There has been little work on the design of the deployment of such multi-source multi-channel live streaming network. We first formulate the problem which comprehensively and realistically captures various cost and delay components and show that it is NP-hard. We then propose an efficient algorithm called COSCOS(Cost Optimization for Multi-source Multi-channel Overlay Streaming) which achieves low deployment cost. Extensive simulation results based on real Internet topologies show that COSCOS outperforms other state-of-the-art schemes by a wide margin, due to its better consideration of the tradeoff between cost and delay.