GB2243267A - Telecommunication - Google Patents

Abstract

A telecommunications equipment such as multiplexer (4) has an internal network-like architecture comprising a highway (64 Kbit/s layer) whereby traffic traverses the equipment and a common medium (message layer) whereby messages internal to the equipment relating to signalling and packet traffic traverse the equipment. At a user interface (5) the respective line card converts the signalling and packet traffic to a common standard and applies it to the message layer, whereas analogue traffic is digitised and applied to the traffic highway. At the output (transmission) interface (7) the messages and traffic are processed and combined as appropriate for the transmission medium (2 Mbit/s layer), which may be optical fibre. The internal network-like architecture of the equipment permits it to be arbitrarily partitioned to suit traffic and geographical requirements e.g. a multiplexer (4) can be physically spread out over a geographical region (Fig 5). <IMAGE>

Description

TELECOMMUNICATIONS This invention relates to telecommunications and in particular to an architecture that is a high level design, for telecommunications equipment such as multiplexers and concentrators.

According to one aspect of the present invention there is provided a telecommunications equipment including a traffic highway whereby traffic traverses the equipment, a common medium whereby messages internal to the equipment relating to signalling and packet traffic traverse the equipment, and means whereby all signalling and packet traffic input to the equipment is converted to messages with a common standard prior to application to the common medium.

According to another aspect of the present invention there is provided a telecommunicatIons equipment comprising a plurality of user interfaces and an external network interface, and having an internal architecture such that there is an internal point-to-point link whereby all signalling and packet traffic traverses the equipment and an internal highway whereby all traffic traverses the equipment, and wherein all input signalling and packet traffic is converted to messages e th an internal standard prior to application to said common medium.

Embodiments of the invention will now be described with reference to the accompanying drawings, in which: Figs la and ib, respectively, illustrate in a schematic form a plane and "side" view of a basic network architecture; Fig 2 illustrates an employment of the basic architecture of Figs la and lb; Fig 3a and Fig 3b illustrate a multiplexer employment of the basic architecture, and Fig 3c illustrates a schematic implementation thereof; Fig 4 illustrates a primary rate multiplexer connected to a higher rate multiplexer employment of the basic architecture, and Fig 5 illustrates a distributed traffic collection employment of the basic architecture.

Recent work on telecommunications network architectures regards networks as comprising linked nodes. Each node either routes or terminates traffic and in addition routes and acts on management messages which control the nodes.

Figs la and lb illustrates in "plan" and "side" view, respectively, a generalised network with an architecture according to the present invention. The network is made up of nodes 1 (solid circles) which route messages. At each node the message may either be for that node or be routed onwards. In telecommunications connections are to be provided and these connections require "ends" and "routes". Shown underneath the message nodes 1 are, therefore, telecommunications traffic nodes 2 (dashed circles) which are either "routers" or "ends". It is assumed for Figs la and lb that there is a larger network to the left of the drawing and the users (subscribers) are to the right of the drawing, so that the network illustrated is an access network.The nodes to the far right of the drawing are, therefore, "ends" and all other nodes are "routers". The star topology is illustrated only as an example, other topologies are equally applicable.

The drawing thus present a simple picture of a network for managing connections. In the access context, user traffic would enter at an "end" at the right hand side, be routed into the larger network (not shown) at the left hand side, traverse the larger network and exit through a similar access network.

In traditional telecommunications systems, the messages which tell the routers how to route come from dialling and are called signalling. In managed access networks for private circuits, the messages come from one or more network management computers and are called network management. Although there may be major differences nn speed requirements, functionally both types of messages accomplish the same things. One of the basic principles behind the architecture of the present invention is to treat both types of messages as identically as possible.

Whereas Fig la shows a plan view of a flat network, the network is actually heirarchical. The heirarchy is based on an analogy with OSI (Open ServIces InterconnectIons) communications with one layer offering a service to a layer above. For instance 34 Mbit/s transmission link offers the service of transporting groups of 2.0e8 Mbit/s channels. The contents of the channels are not operated on, the channels are merely delivered. Similarly a 2.048 Mbit/s channel can be regarded as offering the service of transporting 31 channels each of 64 Kbit/s. This assumes the standard case of achieving that by putting a framing pattern in time slot zero.

Fig ib shows a heirarchy of service (traffic) layers in which, generally, the high bandwidth layers at the bottom offer a service to the layer above. The nodes in Fig lb match those in Fig la. Lower layers may or may not offer the service of carrying messages.

Fig 2 illustrates analogue PSTN traffic entering a network at the sub-rate layer. The traffic is digitised resulting in a 64 Kbit/s rate which is transported in the 64 Kbit/s layer. It should be noted that this layer may have a 2.048 Mbit/s or any other rate, the distinction is that the 64 Kbit/s channel is available for manipulation and routing in this layer.

The signalling is converted into messages and carried by the message layer. Both signalling and traffic pass through a node at which the actual carrying media may change i.e. from a backplane to a cable, for instance.

Once a point (3) is reached where a 2.048 Mbit/s layer is available and where there is no further need for separate signalling and 64 Kbit/s routing, both are converted to use the service of the 2.048 Mbit/s layer, i.e. signalling is converted to the larger network standard in time slot 16 and the traffic is inserted into another time slot. This can now be regarded as a closed container which is routed at points in the 2.048 Mbit/s layer before using the service of a lower layer (e.g. synchronous 155 Mbit/s) for transport across the larger network (not shown).

In the case of PSTN traffic, the process would be reversed up to the 64 Kbit/s layer when a point in the network is reached corresponding to a local exchange. The exchange would route at 64 Kbit/s and the traffic and signalling would descend the layers again for transport across the network.

It should be noted, from Fig 2, that the user signalling messages travelled in the message layer until they were no longer needed for manipulation. They were then transported to a local exchange, not in the message layer but with the traffic. The routing of the 2 Mbit/s "pipes" via lower levels to the local exchange, was controlled by network management messages which were carried, in this example, as a service offered by the synchronous 155 Mbit/s transmission. The reason for this is that networks evolved historically to provide only one service with signalling specific to that service. It should also be noted that routing messages are not the only kind of message in a managed network.

Other kinds exist, such as error monitoring, configuration control etc. These messages co-exist in the message layer and originate and terminate at the appropriate nodes.

The architecture described above is applied in Fig 3a and 3b to a conventional multiplexer applicatIon, such as addressed by PD!X (GB Serial Nos 2185658 and 21a5659), for example. The multiplexer comprises those nodes within dashed box 4, node 5 comprises a user interface with its associated line card. Analogue traffic from a user enters the network at the sub-rate layer as before. The line card itself converts user signalling to messages, which messages conform to an internal standard, such as one based on "Q93i". This is in contrast to a conventional multiplexer which has a central signalling processor dimensioned for the number of circuits supported by the multiplexer e.g. a PDMX has a thirty channel signalling processor. This makes economic sense if the customer wants thirty channels which need signalling conversion.Typical user sites, however, want a mix of user interfaces and are also typically installed with room for growth. Providing processing (a processor) on each line card costs more if the multiplexer is full of that type of line card, otherwise it costs less. Furthermore the use of conversion, on the line cardr to an internal standard permits an easy mix of signalling types and standards. For instance, ISDN and PSTN can co-exist easily, or the network standard can change from CAS to DASS to Q931 without affecting anything other than the network access point.

Continuing to the left in Figs 3a and 3b, the signalling messages and user traffic both pass along a printed backplane (Figs 3a and 3b show a null node 6 in the centre) to the point (7) where the 2.048 Mbit/s layer is accessed. At this point the signalling messages are converted to the appropriate network signalling in time slot 16, user traffic is inserted into the other time slots and the whole is transmitted to the larger network, such as over optical fibre 8.

The message layer may be transported in the network direction as a service offered by the 2.048 bit/s transport (i.e. in an optical overhead or a "stolen" tine slot) or it may be transported separately by a medium such as Ethernet.

Conventional multiplexers usually have a controiler, this reflecting that in reality the message layer is not flat but heirarchical. We propose, however, to make use of the sharp functional partitioning provided by the above architecture, to, for example, make a multiplexer out of user access sub-systems (line cards) and transport (transmission) sub-systems (2 Mbit/s and signalling interface). This permits arbitrary partitioning and packaging to suit traffic and geographical requirements. The architecture will thus permit a multiplexer, for example, to be spread out over a distance (smeared) if necessary, without significant change in the software and hardware building blocks used.

To permit this, the internal messages, i.e.

those transported via the message layer, inside "boxes" such as multiplexer box 4, use a mechanism (common medium) called a multi-master serial bus MMSB for all purposes e.g. signalling, network management, internal control, ISDN packet etc. As is apparent from the above description of a conventional multiplexer using the architecture proposed (Figs 3a and 3b), the signalling is converted at the edges of the box (a product such as a multiplexer) to an internal standard.Assuming the multiplexer is comprised by a single box 4 (Fig 3c), the MMSB 20 can be provided by a suitable conductive track on a back plane with all other elements comprised by user access sub-systems 23 and transmission sub-systems 24 slide-in-units (not shown specifically) engageable with the MSB, and also appropriate bus structures 21 on the back plane for the actual 64 Kbit/s traffic. The slide-in units are intelligent (each includes a processor 22) and the MMSB permits the passing of HDLC messages between the processors of any two units.The use of a standardised message set and the HDLC packet format means that the bus (MMSB) can be replaced by a LAN, WAN or other point-to-point link without any great consequence and hence the component parts of the multiplexer can be geographically spread out without affecting its performance.

Fig 4 shows the architectural case of a conventional primary rate multiplexer 4 (Fig 3a) connected to a larger network (not shown but to the left of Fig 4) via a higher rate multiplexer 9. The higher rate multiplexer 9 offers the services of transporting multiple 2.048 Mbit/s streams and, usually, messages.

In the case of a typical plesiochronous multiplexer, no 2.048 Mbit/s routing function is offered except by physically moving cables. In the case of a sync. mux, routing is available. Higher order and lower order mux functions may, in the future, merge in the access area.

This will then enable users the possibility of both wide and narrow band services, such as voice and LAN interconnect.

Fig 5 illustrates the first stage of spreading out a conventional multiplexer over a geographical region. It is essentially the same as the primary rate multiplexer 4 of Fig 3a. The only difference is that what was previously a null node 6 now routes between the backplane and a transmission link. This is a genuine routing function not a backplane extension and is permitted by the packet based messages used. The block 10 may thus be at the user premises and the block 11 may be at a serving site spaced apart from the user premises although functionally blocks 10 and 11 are equivalent to block 4. There may also be a distance between the serving site 11 and a sync mux 12 with a 2 bit/s transmission link therebetween.The sync mux 12 offers the same service as multiplexer 9 of Fig 4 so that transportation over the sync or pleslochronous layer is offered.

Fig 5 illustrates the purpose of an architecture. Due to differences in scaie, the hardware employed in Figs 4 and 5 will vary at the same nodes.

Since however the architecture is the same, the bulk of the software will also be the same. In addition the architecture confines the variation to specific nodes.

The basic architectural principles employed above may be summarised as follows: 1. Isolation of functions. Everything associated with a particular function, such as a user interface or a network interface, is confined as far as possible inside that function. Everything associated with a user interface is on the respective slide-in-unit.

Everything associated with an upstream interface is in the sub-system that contains that interface. Ideally the latter would also be a single slide-in-unit, but in practice there are applications which would need several units, hence the use of sub-systems which can have a respective processor rather than a processor on each slide-in-unit. In the upstream case, "associated with" functions includes clocks, OSI end system and local environment monitoring as well as the traffic and message carrying service.

2. The sub-systems only have two basic interface types. These are message passing interfaces, for control, error monitoring, signalling etc. and traffic interfaces.

3. In many cases, functions can be combined to minimise costs. In order to achieve maximum re-use of developments and ease future support and modifications, the products should initiaily be structured as though made from separate nodes as illustrated in the architecture. The nodes may then be combined in a single box, slide-in-unit, integrated circuit chip etc., but the functional boundaries should be preserved.

As will be appreciated from the above it is proposed to extend the network architecture into the products used to build the network. Building products with an internal architecture similar to a network architecture permits arbitrary partitioning to suit traffic and geographical requirements. All "messages" internal to the product are carried by one common medium for all purposes i.e. signalling, network management, internal control, ISDN packet etc. Signalling is converted at the "edges" of the product to an internal standard. The actual traffic is carried by a separate medium.

Attention is also directed to our co-pending Application No 8929013.4 (Serial No (D M Goodman - J R Cass 4-3) which describes an equipment practice that the equipment proposed by the present invention may employ.

Claims (9)

CLAIMS:

1. A telecommunications equipment including a traffic highway whereby traffic traverses the equipment, a common medium whereby messages internal to the equipment relating to signalling and packet traffic traverse the equipment, and means whereby all signalling and packet traffic input to the equipment is converted to messages with a common standard prior to application to the common medium.

2. An equipment as claimed in claim 1 and divisible into sub-systems which interconnect with one other via message passing interfaces and traffic interfaces linking the common medium and traffic highway, respectively.

3. An equipment as claimed in claim 1 which is connectible with other said equipments whereby to provide a composite equipment via message passing interfaces and traffic interfaces of the separate equipments which link the common media and traffic highways, respectively, of the equipments to be connected.

4. An equipment as claimed in any one of the preceding claims wherein the common medium is comprised by a conductive track on a back plane, a local area network (LAN), a wide area network (WAN) or another point-to-point link.

5. An equipment as claimed in any one of the preceding claims and including at least one user access sub-system whereby a user accesses the traffic highway and the common medium and a transmission sub-system connected to the traffic highway and the common medium whereby the user is connected to an external network.

6. An equipment as claimed in claim 5 wherein the traffic highway and the common medium are comprised by respective structures on a backplane and the user access and transmission sub-systems are comprised by slide-in units engageable with said backplane structures.

7. An equipment as claimed in claim 5 or claims 6, wherein the sub-systems each include respective processor.

8. A telecommunications equipment comprising a plurality of user interfaces and an external network interface, and having an internal architecture such that there is an internal point-to-point link whereby all signalling and packet traffic traverses the equipment and an internal highway whereby all traffic traverses the equipment, and wherein all input signalling and packet traffic is converted to messages with an internal standard prior to application to said common medium.

9. A telecommunications equipment having an internal network-like architecture and substantially as herein described with reference to and as illustrated in the accompanying drawings.