Energy-Aware Software Video Encoding in Head-End

    Research output: Chapter in Book/Report/Conference proceedingChapter or book articleProfessional

    Abstract

    Today’s extremely fast network connections, as well as powerful processing devices, have enabled high-quality video streaming on the Internet. At the same time, high-definition displays in terminals allow increased video resolution leading also to a higher number of bits to be processed in the head-end. Therefore, demand for improved, novel video encoding technologies is needed. However, green networking ideology drives also for decreasing the carbon footprint in terms of reducing energy and power consumption. Head-end video encoding server usually needs to produce several encoded versions of the same input feed in order to cover different client terminals with unique characteristics, which can lead to extensive energy consumption increasing server-side costs. This work focuses on comparing current video coding format H.264 and state-of-the-art High Efficiency Video Coding (HEVC) for resolving ways to decrease the server-side energy consumption in a software-based video on demand (VoD) encoding. The techniques include selecting a proper hardware and video codec, suitable video resolution and encoding parameters for fulfilling sufficient video quality of experience. The work also introduces comparison between existing video coding technologies. The obtained results indicate that head-end energy consumption can be significantly decreased by using light compression technologies and suitable encoding parameters. In addition, usage of current novel video codecs can slightly increase the head-end energy consumption but obtain clear savings in the network.
    Original languageEnglish
    Title of host publicationGreening Video Distribution Networks
    Subtitle of host publicationEnergy-Efficient Internet Video Delivery
    EditorsAdrian Popescu
    PublisherSpringer
    Chapter4
    Pages81-99
    ISBN (Electronic)978-3-319-71718-0
    ISBN (Print)978-3-319-71717-3
    DOIs
    Publication statusPublished - 31 Jan 2018
    MoE publication typeD2 Article in professional manuals or guides or professional information systems or text book material

    Publication series

    SeriesComputer Communications and Networks
    ISSN1617-7975

    Fingerprint Dive into the research topics of 'Energy-Aware Software Video Encoding in Head-End'. Together they form a unique fingerprint.

    Cite this