IDT Partners

  • history YouTube
  • YouTube Premium
  • Social impact YouTube
  • about 
  • solutions 
  • careers 
  • Adaptive bitrate streaming

    Adaptive bitrate streaming is a technique used in streaming multimedia over computer networks. While in the past most video or audio streaming technologies utilized streaming protocols such as RTP with RTSP, today's adaptive streaming technologies are almost exclusively based on HTTP and designed to work efficiently over large distributed HTTP networks such as the Internet.

    More specifically, and as the implementations in use today are, adaptive bitrate streaming is a method of video streaming over HTTP where the source content is encoded at multiple bit rates, then each of the different bit rate streams are segmented into small multi-second parts. The streaming client is made aware of the available streams at differing bit rates, and segments of the streams by a manifest file. When starting, the client requests the segments from the lowest bit rate stream. If the client finds the download speed is greater than the bit rate of the segment downloaded, then it will request the next higher bit rate segments. Later, if the client finds the download speed for a segment is lower than the bit rate for the segment, and therefore the network throughput has deteriorated, then it will request a lower bit rate segment. The segment size can vary depending on the particular implementation, but they are typically between two (2) and ten (10) seconds.

    Traditional server-driven adaptive bitrate streaming provides consumers of streaming media with the best-possible experience, since the media server automatically adapts to any changes in each user's network and playback conditions. The media and entertainment industry also benefit from adaptive bitrate streaming. As the video space grows, content delivery networks and video providers can provide customers with a superior viewing experience. Adaptive bitrate technology requires additional encoding, but simplifies the overall workflow and creates better results.

    A scalable CDN is used to deliver media streaming to an Internet audience. The CDN receives the stream from the source at its Origin server, then replicates it to many or all of its Edge cache servers. The end-user requests the stream and is redirected to the "closest" Edge server. This can be tested using libdash and the Distributed DASH (D-DASH) dataset, which has several mirrors across Europe, Asia and the US. The use of HTTP-based adaptive streaming allows the Edge server to run a simple HTTP server software, whose licence cost is cheap or free, reducing software licensing cost, compared to costly media server licences (e.g. Adobe Flash Media Streaming Server). The CDN cost for HTTP streaming media is then similar to HTTP web caching CDN cost.

    MPEG-DASH is the only adaptive bit-rate HTTP-based streaming solution that is an international standard MPEG-DASH technology was developed under MPEG. Work on DASH started in 2010; it became a Draft International Standard in January 2011, and an International Standard in November 2011. The MPEG-DASH standard was published as ISO/IEC 23009-1:2012 in April, 2012.

    MPEG-DASH is a technology related to Adobe Systems HTTP Dynamic Streaming, Apple Inc. HTTP Live Streaming (HLS) and Microsoft Smooth Streaming. DASH is based on Adaptive HTTP streaming (AHS) in 3GPP Release 9 and on HTTP Adaptive Streaming (HAS) in Open IPTV Forum Release 2. As part of their collaboration with MPEG, 3GPP Release 10 has adopted DASH (with specific codecs and operating modes) for use over wireless networks.

    "HTTP Dynamic streaming is the process of efficiently delivering streaming video to users by dynamically switching among different streams of varying quality and size during playback. This provides users with the best possible viewing experience their bandwidth and local computer hardware (CPU) can support. Another major goal of dynamic streaming is to make this process smooth and seamless to users, so that if up-scaling or down-scaling the quality of the stream is necessary, it is a smooth and nearly unnoticeable switch without disrupting the continuous playback."

    The latest versions of Flash Player and Flash Media Server support adaptive bit-rate streaming over the traditional RTMP protocol, as well as HTTP, similar to the HTTP-based solutions from Apple and Microsoft, HTTP dynamic streaming being supported in Flash Player 10.1 and later. HTTP-based streaming has the advantage of not requiring any firewall ports being opened outside of the normal ports used by web browsers. HTTP-based streaming also allows video fragments to be cached by browsers, proxies, and CDNs, drastically reducing the load on the source server.

    Smooth Streaming is an IIS Media Services extension that enables adaptive streaming of media to clients over HTTP. The format specification is based on the ISO base media file format and standardized by Microsoft as the Protected Interoperable File Format. Microsoft is actively involved with 3GPP, MPEG and DECE organizations' efforts to standardize adaptive bit-rate HTTP streaming. Microsoft provides Smooth Streaming Client software development kits for Silverlight and Windows Phone 7, as well as a Smooth Streaming Porting Kit that can be used for other client operating systems, such as Apple iOS, Android, and Linux. IIS Media Services 4.0, released in November 2010, introduced a feature which enables Live Smooth Streaming H.264/AAC videos to be dynamically repackaged into the Apple HTTP Adaptive Streaming format and delivered to iOS devices without the need for re-encoding. Microsoft has successfully demonstrated delivery of both live and on-demand 1080p HD video with Smooth Streaming to Silverlight clients. In 2010, Microsoft also partnered with NVIDIA to demonstrate live streaming of 1080p stereoscopic 3D video to PCs equipped with NVIDIA 3D Vision technology.

    In recent years, the benefits of self-learning algorithms in adaptive bitrate streaming have been investigated in academia. While most of the initial self-learning approaches are implemented at the server-side (e.g. performing admission control using reinforcement learning or artificial neural networks), more recent research is focusing on the development of self-learning HTTP Adaptive Streaming clients. Multiple approaches have been presented in literature using the SARSA or Q-learning algorithm. In all of these approaches, the client state is modeled using, among others, information about the current perceived network throughput and buffer filling level. Based on this information, the self-learning client autonomously decides which quality level to select for the next video segment. The learning process is steered using feedback information, representing the Quality of Experience (QoE) (e.g. based on the quality level, the number of switches and the number of video freezes). Furthermore, it was shown that multi-agent Q-learning can be applied to improve QoE fairness among multiple adaptive streaming clients.