4.2 End-to-End System Architecture

A streaming system is a real-time system of the nonconversational type. It is real-time because the playback of continuous media, such as audio and video, must occur in an isochronous fashion. A streaming application is different from a conversational application because it has the following properties:

One-way data distribution. The media flow is always unidirectional, from the streaming server to the mobile client (in the downlink direction). Normally, the user has limited control over a streaming session, and there is not a high level of interactivity between mobile client and streaming server. Typical user control commands in the uplink direction include PLAY, PAUSE, STOP, FAST-FORWARD, and REWIND.
Offline media encoding. A streaming system is similar to a Video On Demand system, where the user can play only prestored content. This content is not encoded in real-time, but in an offline fashion using specific content creation tools.
Not highly delay sensitive. Because high interactivity is not a requirement of a streaming system, end-to-end delays can be relaxed. For example, the time required by the streaming client to execute a command issued by the user (such as PLAY) does not need to be on the order of milliseconds. Media can be streamed after an initial latency period. This allows the mobile client to smooth out eventual network jitter without compromising user QoS.

Figure 4.1 describes the high-level architecture of a typical mobile multimedia streaming system over an IP-based mobile network. We will follow an end-to-end approach, analyzing the system in its different parts. A mobile streaming system consists mainly of three components: (1) the streaming server, (2) the mobile network, and (3) the mobile streaming client.

click to expand
Figure 4.1: A typical mobile multimedia streaming system.

The streaming server is connected to a fixed IP network and can reside either within the mobile operator's domain or outside it (e.g., the Internet). The location of the streaming server is important when considering the end-to-end quality of service of a streaming service. In fact, if the server is located in the public Internet, the QoS of the network trunk between the streaming server and the mobile network is not usually controlled by the mobile operator, and it can be of the best-effort type in the worst case. This may have impact on the perceived streaming service user quality.

The content created offline is loaded onto the streaming server before a user can actually request its playback. The content that is estimated to be highly requested can be replicated or cached in proxy servers using appropriate techniques that make use of usage patterns.

The mobile network carries multimedia streaming traffic mainly between the streaming server and the mobile streaming client. A logical connection established between the network and the mobile client addresses is called PDP (Packet Data Protocol) context. This uses physical transport channels in the downlink and uplink directions to enable the data transfer in the two directions.

The mobile streaming client keeps a radio connection with the mobile network, utilizing the allocated PDP context for data transfer. The mobile client has the possibility to roam (i.e., upon mobility, change the network operator without affecting the received service), provided there is always radio coverage to guarantee the service. The data flows received by the mobile client in the downlink direction are, for example, audio and video plus additional information for session establishment, control, and media synchronization. The data flows sent by the client in the uplink direction are mainly session control data and QoS reports. The streaming server may react accordingly upon reception of the QoS reports, taking appropriate actions for guaranteeing the best-possible media quality at any instant.