DASH is based on a clever way of structuring the content that allows a great adaptability during the streaming without requiring any server side computation.
Periods are used to delimit content depending on the time. It can be used to delimit chapters, or to add advertisements that occur at the beginning, during or at the end of a video.
The representation level is the level DASH uses to offer the same content at different levels of resolution.
For example, a adaptation set containing images have a representation for each available resolution (it might be 480p, 720p, 1080p, etc\ldots).
This allows a user to choose its representation and change it during the video, but most importantly, since the software is able to estimate its downloading speed based on the time it took to download data in the past, it is able to find the optimal resolution, being the highest resolution that arrives on time to avoid stalling.
For example, a representation of images of a chapter of a movie can be heavy and long to download.
However, downloading heavy files is not suitable for streaming because it prevents the dynamicity of it: if the user requests to change the level of resolution of a video, the system would either have to wait until the file is totally downloaded, or cancel the request, making all the progress done unusable.
Segments are used to prevent this behaviour. They typically encode files that last approximately one second of video, and give the software a great ability to dynamically adapt to the system. If a user wants to seek somewhere else in the video, only one second of data can be lost, and only one second of data has to be downloaded for the playback to resume.
\subsubsection{Client side computation}
Once a video is encoded in DASH format, once the files have been structured and the MPD has been generated, they can simply be put on a static HTTP server that does no computation other than serving files when it receives requests.
All the intelligence and the decision making is moved to the client side.
A client typically starts by downloading the MPD file, and then proceeds on downloading segments of the different adaptation sets that he needs, estimating itself its downloading speed and choosing itself whether it needs to change representation or not.
DASH-SRD (Spatial Relationship Description,~\cite{dash-srd}) is a feature that extends the DASH standard to allow stream only a spatial subpart of a video to a device.
It works by encoding a video at multiple resolutions, and tiling the highest resolutions, that way, a client can choose to download either the low resolution of the whole video or higher resolutions of a subpart of the video (see Figure~\ref{sota:srd-png}).
For each tile of the video, an adaptation set is declared in the MPD, and a supplemental property is defined in order to give the client information about the file.
This supplemental property contains many elements, but the most important ones are the position ($x$ and $y$) and the size (width and height) of the tile describing the position of the tile in relation to the full video. An example of such a property is given in Listing~\ref{sota:srd-xml}.
Essentially, this feature is a way of achieving view-dependent streaming, since the client only displays a part of the video and can avoid downloading content that will not be displayed.
This is especially interesting in the context of 3D streaming since we have this same pattern of a user viewing only a part of a content.
We briefly survey other research on prefetching that focuses on non-continuous interaction in other types of media.
In the context of navigating in a video, a recent work by carlier et al.~\cite{video-bookmarks} prefetches video chunks located after bookmarks along the video timeline.
Their work, however, focuses on changing the user behavior to improve the prefetching hit rate, by depicting the state of the prefetched buffer to the user.
Carlier et al.\ also consider prefetching in the context of zoomable videos in an earlier work~\cite{zoomable-video}, and showed that predicting which region of videos the user will zoom into or pan to by analyzing interaction traces from users is difficult.
Prefetching for navigation through a sequence of short online videos is considered by khemmarat et al in~\cite{user-generated-videos}.
Each switch from the current video to the next can be treated as a non-continuous interaction.
The authors proposed recommendation-aware prefetching --- to prefetch the prefix of videos from the search result list and related video list, as these videos are likely to be of interest to the user and other users from the same community.
Grigoras et al.~\cite{video-navigation-mpd} consider the problem of prefetching in the context of a hypervideo; non-continuous interaction happens when users click on a hyperlink in the video.
They propose a formal framework that captures the click probability, the bandwidth, and the bit rate of videos as a markov decision problem, and derive an optimal prefetching policy.
Zhao and Ooi~\cite{joserlin} propose joserlin, a generic framework for prefetching that applies to any non-continuous media, but focuses on peer-to-peer streaming applications.
They do not predict which item to prefetch, but rather focus on how to schedule the prefetch request and response.
There is a huge body of work on prefetching web objects in the context of the world wide web. Interested readers can refer to numerous surveys written on this topic (e.g.,~\cite{survey-caching-prefetching}).
