Content Delivery Using Replicated Digital Fountains

With a majority of Internet traffic being predicted to be caused by content delivery, it is clear that content delivery applications will consume much of the resources on the Internet. This paper considers the problem of cost-efficient content delivery, in which the application incurs both a network delivery cost (e.g., from cross ISP traffic or, more generally, operation/energy costs at Internet routers) and costs at the servers (e.g., due to cost of ownership, energy, or disk bandwidth). While the cost objective and the absolute cost tradeoff may be different from case to case, we argue that an architecture with distributed servers, each using digital fountain delivery, may be an attractive candidate architecture when considering the total content delivery cost. Within the context of a simple system model, we then determine optimal server selection policies for such an architecture, and derive analytic expressions for their associated delivery costs. A readily-implementable heuristic policy is proposed that is found to achieve within 10% of the minimal cost. Finally, we show how our results for content download can also be applied to streaming video delivery.