An optimized distributed video-on-demand streaming system: Theory and design

Abstract

We present a general framework for a distributed VoD content distribution problem by formulating an optimization problem yielding a highly distributed implementation that is highly scalable and resilient to changes in demand. Our solution takes into account several individual node resource constraints including disk space, network link bandwidth, and node-I/O degree bound. First, we present a natural formulation that is NP-hard. Next, we design a simple fractional storage architecture based on codes to “fluidify” the content, thereby yielding a convex content placement problem. Third, we use a recently developed Markov approximation technique to solve the NP-hard problem of topology selection under node degree bound, and propose a simple distributed solution. We prove analytically that our algorithm achieves close-to-optimal performance. We establish via simulations that the system is robust to changes in user demand or in network condition and churn.