A media service network is a network through which media services are provided to an end user. The services are implemented by applications controlling media resource functions on a media server. In this section we discuss the components that make up a media service network, illustrate three representative communication network configurations into which media servers are deployed, and describe the interfaces used by media service applications.

There are a number of architectures in use today that define the composition of a media service network. Some are specified by industry standards bodies and others are proprietary. Most can be generalized to a simple model where the media server interoperates with a number of other components to form the complete solution.

Components of a Media Service Network

We will first define the basic components of a media service network: end-user terminals, application server, media server, media store, and gateways.

End users access media services through the public network from a local device we will call a terminal. Examples of terminals are a plain old telephone (POT), a video-enabled cell phone, and a multimedia PC.

The application logic that realizes the media service is hosted by an application server. The application server provides the runtime execution environment for the application.

Media processing such as transcoding and video image processing (e.g., text overlay, resizing) are performed by media processing resources on the media server.

The media store is responsible for the storage and retrieval of the media (i.e., to/from disk).

If the media server is interoperating with disparate networks (where multiple transport and communications protocols are in use), the media server will have to interoperate with gateways that are responsible for translating between the different network protocols. Gateways are typically classified as signaling gateways that translate one signaling protocol to another and media gateways that translate from one media format to another. We explain signaling in the following section.

Figure 1: Basic media service network architecture
click image for larger view
 

Control and Data Plane

Components in a media service network can be thought of as interoperating in two planes. The data plane, where the actual media traffic flows, is depicted as the thick lines in Figure 1. The control plane, shown as the thin lines, is where control and signaling messages flow. Signaling messages are used to establish and manage the media session between the end-user terminal and the media server. The session contains the state and any other context associated with the connection. The terms "session" and "call," as in a telephone call, are used interchangeably. Signaling typically originates from the end-user terminal and the application, potentially being translated by a signaling gateway. As mentioned, the media traffic flows along the data plane. Routing and flow of the media are controlled using a connection control protocol in the control plane. Connection control typically originates from the application and is directed toward the terminal or gateway and media server.

The application logic controls the media server through a media control API, discussed in more detail later.

Composing a Media Server

The media server is the architectural element responsible for transmitting media content to an end user over a communications network. The content may be in the form of video, audio, text, or a combination of the three. The media server often contains digital signal processing resources to process or transform the media prior to transmitting it to the user (e.g., gain and speed control of an audio stream). In addition to transmitting media content to the end user, a media server may also be capable of accepting control input from the end user's terminal. This input may be used by the media service application, making it interactive.

Figure 1 illustrates the logical components of a media service network. In this diagram the application server, media server, media store, and gateway are shown as separate entities. This is a logical organization and does not necessarily imply a physical decomposition. In practice, depending on the individual system requirements, any or all of these elements can be combined within a single node. In the following sections we will often simplify the diagrams by combining the application server, media server, and media store elements into a single node.