Cross-layer signalling support for adaptive video encoding and transmission

Cyril Bergeron, Lorenzo Iacobelli, Esa Piri, Janne Vehkaperä

    Research output: Contribution to conferenceConference PosterScientific

    Abstract

    Following the joint source and channel coding paradigm, we propose an application controlling strategy to allow fast adaptation of multimedia transmission to the impairments of an IP wireless network by using a new cross-layer signalling architecture. To this purpose, a Rate-Distortion Optimization (RDO) algorithm based on hybrid Cauchy-Laplace distribution is presented. In order to support the adaptation and to enable the application controlling strategy by providing the necessary signalling information, a flexible and cloud computing based cross-layer event delivery entity is proposed. This entity provides a unified signalling architecture for accessing such information that cannot be accessed by using existing controlling mechanisms and which is necessary for the adaptation of the multimedia transmission. Information obtained through this event delivery entity is used by the RDO algorithm to define the parameters providing the best perceived quality at the receiver side in terms both of compression and optionally of error correcting redundancy insertion to minimize the error impact with a GOP adaptation rate. This technique is demonstrated here for H.264 streams.
    Original languageEnglish
    Publication statusPublished - 2015
    Event24th European Conference on Networks and Communications, EuCNC 2015 - Paris, France
    Duration: 29 Jun 20152 Jul 2015
    Conference number: 24

    Conference

    Conference24th European Conference on Networks and Communications, EuCNC 2015
    Abbreviated titleEuCNC 2015
    Country/TerritoryFrance
    CityParis
    Period29/06/152/07/15

    Keywords

    • rate-distortion optimization
    • H.264
    • cross-layer signalling
    • DDE
    • error prone channel
    • multimedia adaptation

    Fingerprint

    Dive into the research topics of 'Cross-layer signalling support for adaptive video encoding and transmission'. Together they form a unique fingerprint.

    Cite this