MPEG DASH provides formats that are suitable to stream segmented media content over HTTP.
DASH clients follow a client-pull paradigm by adapting their requests based on the available bandwidth and other local resources.
This has proven to be easier to deploy over CDN infrastructure than server-push technologies.
However, this decentralised nature introduces new challenges such as offering a consistent and higher quality of service for premium users.
MPEG addressed this issue in the nrecent MPEG DASH part 5, Server and Network-assisted DASH (SAND).
The key features of SAND are asynchronous network-to-client and network-to-network communication of quality-related assisting information.
In addition, DASH-IF is further defining interoperable guidelines to optimise SAND deployments in a variety of environments: home network, over-the-top etc.
Streaming media technologies have evolved over the past few years. HTTP-based adaptive streaming is today the technology of choice for streaming over the Internet.
Many standards and industry bodies, including DVB, 3GPP, HbbTV and DASH-IF, have adopted the Dynamic Adaptive Streaming over HTTP (DASH) standard (1) published in 2012 by MPEG.
The DASH formats are designed to be used in client-pull based deployments primarily with HTTP protocol for media delivery.
A client first retrieves a manifest file, a Media Presentation Description (MPD), and then it selects, retrieves and renders content segments based on that metadata.
By leveraging HTTP communication, DASH offers some fundamental benefits over the other streaming technologies such as: firewall pass through, content being stored on regular HTTP servers, or high scalability.
Most importantly, the DASH client selects segments of quality and size that fits within the bandwidth it has available.
This way, buffer underruns are prevented and the end user can benefit from a continuous media experience.
However, the fundamental decentralised and client-driven nature of DASH also has some drawbacks.
Service providers may not necessarily have control over the client behaviour. Consequently, it may be difficult to offer a consistent or a premium quality of service.
There are many examples of situations where quality of experience can be affected.
Resources announced in the MPD may become outdated after a network failure or reconfiguration, resulting in misdirected and unsuccessful DASH segment requests by the client.
A DASH client can mistakenly switch to lower quality segments when a mobile hand-over or cache miss is interpreted as a bandwidth reduction.
Massive live DASH streaming may lead to cascades of cache misses in CDNs.
A DASH client may unnecessarily start a stream with lower quality segments, and only ramp up after it has obtained bandwidth information based on a number of initial segments.
Multiple DASH clients may compete for a shared bandwidth, leading to unwanted mutual interactions and possibly oscillations.
As a consequence , service providers may not be able to guarantee a premium quality of service with DASH, even in managed networks where regular DASH clients may not fully take advantage of the offered quality of service features.
In 2013, MPEG started the Core Experiment on Server and Network-assisted DASH (CE-SAND), in which experts collected use cases and explored solutions.
Based on the results of CE-SAND, MPEG initiated in February 2015 the edition of MPEF DASH part 5 SAND, ISE/IEC 23009-5 (2).
This new part of MPEG DASH defines and architecture, data models and a protocol to solve the issues mentioned above.
MPEG DASH SAND is expected to be published in Q3 2016.
The paper presents the SAND architecture as defined in the MPEG DASH part 5 in terms of terminologies, data formats and protocols.
In addition, concrete deployment scenarios are presented based on the ongoing SAND discussions in DASH-IF wherein experts are establishing guidelines to deploy DASH services augmented by SAND.
The scenarios highlight the appropriate SAND messages to use as well as the achieved benefits for both the DASH clients and the service providers.
DOWNLOAD THE FULL TECH PAPER BELOW
- PDF, Size 0.25 mb