For satellite transmission and to remain consistent with already existing MPEG-2 technology6, MPEG-4 TSs (or other data) are further encapsulated in Multiprotocol Encapsulation (MPE – RFC 3016) and then segmented again and placed into TS streams via a device called IP Encapsulator (IPE; Fig. A4.5). MPE is used to transmit datagrams that exceed the length of the DVB “cell,” just as Asynchronous Transfer Mode Adaptation Layer 5 (AAL5) is used for a
similar function in an ATM context. MPE allows one to encapsulate IP packets into MPEG-2 TSs (“packets,” or “cells”; Fig. A4.6). IPEs handle statistical multiplexing and facilitate coexistence. IPE receives IP packets from an Ethernet connection and encapsulates packets using MPE, and then maps these streams into an MPEG-2 TS. Once the device has encapsulated the data, the IPE forwards the data packets to a satellite link. Generic data (IP) for transmission over the MPEG-2 transport multiplex (or IP packets containing MPEG-4 video) is passed to an encapsulator that typically receives PDUs (Ethernet frames, IP datagrams, or other network layer packets); the encapsulator formats each PDU into a series of TS packets (usually after adding an encapsulation header) that are sent over a TS logical channel. The MPE packet has the format shown in Fig. A4.7. Figure A4.8 shows the encapsulation process.
Note: IPEs are usually not employed if the output of the layer 2 switch is connected to a router for transmission over a terrestrial network; in this case, the headend is responsible for proper downstream enveloping and distribution of the traffic to the ultimate consumer. In other pure, IP-based video environments, where DVB-S or DVB-S2 are not used (e.g., a greenfield IP network that is designed to just handle video), the TSs are included in IP packets that are then transmitted as needed (Fig. A4.9). Specifically, with the current generation of equipment, the encoder will typically generate IP packets; these have a source IP address and a single or multicast IP address. The advantage of having video in IP format is that it can be carried over a regular (pure) Local Area Network (LAN) or carrier Wide Area Network (WAN).
Consider Fig. A4.9 again. It clearly depicts video (and audio) information being organized into PESs that are then segmented into TSs. Examining the protocol stack of Fig. A4.9 one should note that in a traditional MPEG-2 environment of DTV, either over-the-air transmission or cable TV transmission, the TSs are handled directly by an MPEG-2-ready infrastructure formally known as an MPEG-2 Transport Multiplex (see left-hand side stack). As explained, MPEG-2 Transport Multiplex offers a number of parallel channels that are known as TS logical channels. Each TS logical channel is uniquely identified by the PID value that is carried in the header of each MPEG-2 TS packet. TS logical channels are independently numbered on each MPEG-2 TS multiplex (MUX). As just noted, the service provided by an MPEG-2 transport multiplex offers a number of parallel channels that correspond to logical links (forming the MPEG TS) [24, 28]. The MPEG-2 TS has been widely accepted not only for providing digital TV services but also as a subnetwork technology for building IP networks, say in cable TV–based Internet access.
There may be an interest in also carrying actual IP datagrams over this MPEG-2 transport multiplex infrastructure (this may be generic IP data or IP packets emerging from MPEG-4 encoders that contain 7 MPEG-4 frames.) To handle this requirement, packet data for transmission over an MPEG-2 transport multiplex is passed to an IPE. This receives PDUs, such as Ethernet frames or IP packets, and formats each into a Subnetwork Data Unit (SNDU), by adding an encapsulation header and trailer. The SNDUs are subsequently fragmented into a series of TS packets. To receive IP packets over an MPEG-2 TS Multiplex, a receiver needs to identify the specific TS multiplex (physical link) and also the TS logical channel (the PID value of a logical link). It is common for a number of MPEG-2 TS logical channels to carry SNDUs; therefore, a receiver must filter (accept) IP packets sent with a number of PID values, and must independently reassemble each SNDU [29]. Some applications require transmission of MPEG-4 streams over a preexisting MPEG-2 infrastructure, for example, in a cable TV application. This is also done via the IPE; here the IP packets generated by the MPEG-4 encoder are considered (treated) as if they were data, as just described above in this paragraph (Fig. A4.10).
The encapsulator receives PDUs (e.g., IP packets or Ethernet frames) and formats these into SNDUs. An encapsulation (or convergence) protocol transports each SNDU over the MPEG-2 TS service and provides the appropriate mechanisms to deliver the encapsulated PDU to the receiver IP interface. In forming an SNDU, the encapsulation protocol typically adds header fields that carry protocol control information, such as the length of SNDU, receiver address, multiplexing information, payload type, and sequence numbers. The SNDU payload is typically followed by a trailer that carries an Integrity Check (e.g., Cyclic Redundancy Check, CRC). When required, an SNDU may be fragmented across a number of TS packets (Figs A4.11 and A4.12) [29].
In summary, the standard DVB way of carrying IP datagrams in an MPEG-2 TS is to use MPE; with MPE each IP datagram is encapsulated into one MPE section. A stream of MPE sections are then put into an ES, that is, a stream of MPEG-2 TS packets with a particular PID. Each MPE section has a 12-B header, a 4-B CRC (CRC-32) tail, and a payload length, which is identical to the length of the IP datagram that is carried by the MPE section [30].