Network-friendly box-powered video delivery system

On-demand video delivery with a content delivery network (CDN) solely based on set-top-boxes (STB) has been introduced recently. This architecture combines the load balancing and fault tolerating features of peer-to-peer systems with stableness of a server based CDN, by storing contents in STBs. Since STBs usually have much longer online time than traditional PC based peers, they are ideal for on-demand video services. Currently, videos are separated into pieces and randomly stored in STBs. The video requests from clients are redirected to the nearest boxes. However, random strategy may lead to some costly and inefficient allocations, for example, a video part could be placed much further than other parts of the same video, while two close boxes host the same part. This paper aims to reduce the downloading cost by exploiting the network location of boxes when allocating the videos. We show that optimizing the allocation is NP-hard. Two locality-based online heuristics supporting transient boxes are proposed. The simulations with realistic network settings demonstrate that our heuristics have comparable performance to an existing approximate algorithm and outperform the random allocation.