IEC 62379
Common Control Interface
for
networked audio and video equipment
Most current networks use IP (v4 or v6), but there is a growing recognition that this form of packet switching has significant shortcomings for the "streamed" media which form an increasing proportion of the traffic carried on the Internet. There have been a number of "clean slate" projects such as the 100x100 project investigating more appropriate networking systems.
Until recently it was thought that in the "real world" it was necessary to use the same routing techniques as the current Internet, and that "next generation" IP networks would provide a "good enough" service. Now, however, work is beginning on standardising telecommunications technologies that are not restricted to IP-style packet switching. Groups working on this topic include:
ETSI ISG NGP, which is looking at technologies that would be suitable for next-generation Radio Access Networks, connecting mobile devices to the Internet and other services; and
ISO/IEC JTC1/SC6/WG7, which has produced a Technical Report on the problems with IP and is now developing standards (ISO 21558 Architecture and ISO 21559 Protocols and mechanisms) for the new network technology.
When these networks are deployed they will need to co-exist and interwork with the previous technologies, for instance via gateways. To give the best performance, applications will need to be able to take advantage of improvements in Quality of Service (QoS) parameters such as latency and jitter, which are likely to be significantly lower on the new networks, and of new services which they may provide.
Some likely features of the new networks, and the way in which the protocols specified in IEC 62379 -- and particularly Part 5-2 -- support them, are described below.
Future networks are likely to support circuit switching, in which routing decisions (which can be computationally intensive) are made by software before transmission of the data begins, and data forwarding (which, once the routing decisions have been made, is very simple) is handled by dedicated hardware. This reduces complexity in routers, and therefore also reduces the amount of power they consume.
IP networks also separate discovery of the destination address, though not of the route to that address, from forwarding of the data, with protocols such as DNS and SIP. Also, MPLS can provide predefined routing within parts of an IP network.
The signalling protocols specified in IEC 62379-5-2 provide for discovery of the route before transmission of the data begins, as in the call connection procedures that are used on circuit-switched networks, but the definition of a route (or rather, of a link, which may be a whole subnetwork) is broad enough that this can also apply to the discovery of the destination address on a IP network.
The signalling protocols allow alternative routes to be discovered, and data flow to be switched between routes. This enables both failover when a link fails in a static network and handover when a mobile device moves from one cell or point of attachment to another.
The level of resilience can be chosen to suit the application, for instance in the case of failover the new route may be found when the failure occurs, or a backup route may be chosen and kept in reserve until needed, or multiple copies of the data may be sent by different routes.
No assumptions are made about the form of addressing used by the underlying network. The signalling messages do not have a fixed format (or fixed length field) for an address, so a very wide range of addresses can be used, including URLs.
An address may include an identifier for a piece of equipment or a service, or a locator which shows where to find it, or both. There may be more than one locator, and the locators are then applied sequentially, in the same way that in a phone number containing a country code and area code, the call is routed first to the country and then to the area within that country.
Provision is made for the application, or the source of the data stream, to specify the bandwidth required. This information can then be used to reserve capacity on the links traversed by the data, if this is supported by the underlying network.
The signalling messages report the end-to-end QoS to be expected. This allows an application to know, for example, whether to expect the incoming data packets to arrive at regular or irregular intervals, and allocate buffer space appropriately, to minimise latency in the first case and dropped packets in the second.
The signalling messages also describe other features of the data, such as encoding and encapsulation, so they contain all the information the recipient will need to decode the data. This can be used in a variety of ways, including negotiating the best compromise between quality of compressed material and capacity available in the network, also to support transcoding in gateways. The way in which formats and encapsulations are represented is fully extensible, providing a globally-unique coding not only for standard formats and protocols but also for those that are manufacturer- or application-specific or experimental.
Additional information that may be carried by the route discovery protocols includes charging information. This allows the equivalent of premium rate telephone calls to be provided.