A Synergic Architecture for Content Distribution in Integrated Satellite and Terrestrial Networks

2020 
Satellites are attracting increasing attention as novel broadband Internet access. It is a trend for satellites to enhance the content distribution efficiency in cooperation with the terrestrial network. Current studies mainly treat Low Earth Orbit (LEO) satellites as stable overlays by a series of snapshots, ignoring satellites' capability of cache, movement and networking. To further improve the global content distribution efficiency in the integrated satellite and terrestrial networks, this paper proposes a synergic distribution architecture combining satellites with the Content Delivery Network (CDN). The architecture takes advantage of LEO satellites' mobility characteristic to cache and deliver multiple times along the trajectory. The content distribution process not only involves content transfer among static nodes in the terrestrial CDN, but also mobile satellites serving as couriers. To optimize the overall bandwidth saving and reduce the distribution time, the distribution process is modeled as a maximum matching problem between target receivers and satellites' trajectory. The problem is solved with an integer linear programming. The distribution architecture is analyzed with the edge servers distribution of a CDN provider, Akamai, and an emerging satellite constellation, Starlink. The simulation results show that the proposed architecture improves the distribution efficiency of existing CDNs with significant savings of bandwidth and delivery time. A single satellite in a cache-and-multiple-deliver manner saves tens of TBs bandwidth consumption in one revolution. And the satellite network saves 25% (40%) of bandwidth consumption in 30 (60) minutes when satellites multicast at a bottleneck rate on the ground. The distribution time is reduced by 25% in half an hour at the same time.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    12
    References
    0
    Citations
    NaN
    KQI
    []
    Baidu
    map