FIELD OF THE INVENTION
The invention relates to telecommunication exchanges, methods of using such exchanges, switching systems for communication networks, methods of using such systems, and methods of routing traffic in networks.
BACKGROUND OF THE INVENTION
Known switched telecommunication networks, in general, consist of a hierarchical structure as shown in FIG. 1, and comprise some of the following set of elements:
trunk routes and trunk circuits
Local exchanges are nodes in the network to which subscribers are connected. The low calling rates generated by subscribers are concentrated into streams of traffic which are routed onwards to trunk exchanges.
Trunk exchanges do not have subscribers connected to them. They operate on traffic which has been received from other exchanges and which has been pre-concentrated by the local exchanges.
Between any two exchanges (of any type), there exists a trunk route; each trunk route consists of a set of trunk circuits. Each trunk circuit, at any one time, carries the voice (or data) samples for one telephone call.
Trunk exchanges can act as either transit exchanges or tandem exchanges, depending upon their position in the network. It is also possible for an exchange to be a hybrid; for example, a local exchange can include some tandem functions. Occasionally, subscribers are connected directly to a trunk exchange; this can be in order to offer special services or for special reasons of topology or efficiency. The location and type of exchanges is a balance between transmission and switching costs, and the availability of sites.
Specialised trunk networks may also be employed to offer services where such services are not available on the main trunk exchanges or where significantly different calling patterns apply. A good example of such a specialised network is BT's Digital Derived Services Network which offers freephone 0800 and local rate 0345 services. Specialised networks are used when the traffic volumes make it difficult to justify the expense of upgrading the main trunk network or where service features are offered by a different vendor. However, the cost of switching through both trunk exchanges and specialised exchanges may be high if traffic starts to rise significantly. In some cases, the services will then be added in to the main trunk exchanges, and the specialised exchanges can be developed further or re-engineered to provide other services.
Signalling System No.7 (SS No.7) is based upon ITU-T Recommendations Q701-706, Q721-725 and Q761-767. SS No.7 is designed for digital communication networks, including ISDN, which use stored program control exchanges. SS No.7 is optimised for 64 kbit/s bearers, however it can be used on 56 kbit/s bearers and can be operated at lower rates over analogue bearers (e.g. 4.8 kbit/s). It is currently widely used in the UK to set up and clear down telephone calls over 64 kbit/s trunk circuits between digital exchanges. SS No.7 is constructed on a modular basis. SS No.7 (in a narrowband environment) consists of a four-level hierarchy in two main parts: the message transfer part (MTP) which covers levels 1 to 3, and the application-dependent user part (e.g. telephone user part or ISDN user part) which is at level 4. FIG. 2 is a schematic of the overall SS No.7 structure in which the four levels correspond approximately to the first four levels of the well known seven-layer OSI model.
SS No.7 can also be used in a broadband environment. In such a situation, the broadband version of ISUP (commonly known as B-ISUP) would be used (based upon ITU-T Recommendations Q1761-1764) together with the same MTP-L3 as is used for narrowband. At the levels below MTP-L3, ATM would be used; with, for example, the ATM SAAL (based upon ITU-T Recommendations Q2100, etc.) being used instead of MTP-L2.
A brief summary pertaining to SS No.7 in a narrowband environment follows.
SS No.7 conveys information within signal units (SU), which are delimited by flags. There are three types of signal units, namely FISUs, LSSUs and MSUs. MSUs are the type of signal units which carry the signalling messages which are generated or handled by level 3, level 4 and above.
Level 1 (MTP-L1) defines the physical and functional characteristics of the signalling data link and the access to it. The signalling data link is a full duplex link which can operate over terrestrial or satellite channels. Normally a signalling data link is TS16 in the 32 time-slot 2048 kbit/s PCM system; however TS67-70 in the higher-order 8 Mbit/s multiplex can also be used; the signalling bit rate being 64 kbit/s in both cases. Lower rate signalling data links can also be used.
Level 2 (MTP-L2) carries out the signalling link functions. These include the adaptation between the processor's signal units and the continuous bit stream which is used over the signalling data link (including the insertion and removal of flags plus bit stuffing). Level 2 deals with alignment, with error detection and correction (using cyclic redundancy code x16+x12+x5+1), and with the monitoring of the bit error rate. Where necessary, level 2 generates and receives LSSUs and FISUs. Level 2 transfers MSUs to and from level 3.
Level 3 (MTP-L3) carries out the signalling message handling functions; that is, it directs signalling messages to the correct signalling link or user part. Level 3 also carries out signalling network management functions; that is, it reconfigures signalling links after a failure or recovery, and generates and receives MTP-L3 network management signalling messages.
Level 4 is the user part; for example, it can be the TUP (telephone user part) or ISUP (ISDN user part).
All signal units (SU) include the following fields:
| || |
| || |
| ||F = ||Flag (01111110) |
| ||CK = ||Check bits |
| ||LI = ||Length indicator |
| ||FIB = ||Forward indicator bit |
| ||FSN = ||Forward sequence number (this is achieved by |
| || ||incrementing the previous FSN by 1 in modulo |
| || ||128) |
| ||BIB = ||Backward indicator bit |
| ||BSN = ||Backward sequence number |
| || |
In addition, MSUs include the following fields:
| || |
| || |
| ||SIO = ||Service information octet |
| ||SIF = ||Signalling information field. |
| || |
The service information octet and the signalling information field constitute the real contents of the signalling message as seen by the user parts and above. The service information octet identifies the type of user part. The signalling information field contains a routing label and the information content of the signalling message. The routing label includes the originating point code and the destination point code. (Note that the routing label for broadband SS No.7 is exactly the same as for narrowband SS No.7.)
The routing of messages such as MSUs, either for controlling telephone calls, for managing the signalling network, or for other purposes (e.g. accessing special services or data-bases), requires that the signalling points in the signalling network are given unique addresses. With SS No.7, these unique addresses are called point codes. In the SS No.7 signalling network, telephone exchanges are examples of signalling points.
The originating point code and the destination point code in a signalling message identify the sender and receiver of that signalling message. Thus this pair of point codes identify the signalling route set between the sending signalling point and the receiving signalling point. The signalling route set consists of all possible signalling paths between those two signalling points.
If the user part associated with the signalling message is the TUP or ISUP (or B-ISUP), then the pair of point codes within the signalling message identifies the sending exchange and the receiving exchange, and identifies the signalling route set between those two exchanges. This same pair of point codes also identifies the trunk route between those two exchanges (used by the telephone call with which the signalling message is associated).
In the situation where associated SS No.7 signalling is used, then the signalling route set between the sending exchange and the receiving exchange follows the same (direct) physical path as the trunk route.
In the situation where quasi-associated SS No.7 signalling is used, the signalling route set may follow a different physical path from the trunk route. In this case, the signalling route set will traverse one or more SS No.7 signalling transfer points. However, the signalling route set is still between the sending and the receiving exchange and there is still a one-to-one correspondence between the trunk route and its corresponding signalling route set. Note that, when quasi-associated signalling is used, it is possible to have a signalling network which is separate from the voice network.
A signalling link set is defined as the set of all the signalling links between two adjacent signalling points. In the case of associated SS No.7 signalling, the signalling route set between a sending exchange and a receiving exchange is carried via the signalling link set between those two exchanges. In the case of quasi-associated SS No.7 signalling, the signalling route set between the sending and receiving exchanges will be carried by a signalling link set between the sending exchange and the signalling transfer point and then carried by a (different) signalling link set between the signalling transfer point and the receiving exchange.
The allocation of these unique point codes to signalling points (e.g. telephone exchanges) is centrally controlled for each network on a country by country basis. A limited number of bits are available in the SS No.7 standards for defining point codes. Therefore unlimited expansion is not possible within the standards.
An exception to the rule of one point code per exchange occurs for international exchanges, which may have one point code in a country's national network and another in the international network. Similarly, gateway exchanges can have multiple point codes, each associated with a different network. When quasi-associated signalling is the norm in a network, there can exist gateway signalling transfer points, and such signalling transfer points can have multiple point codes; one particular example of this is described in U.S. Pat. No. 5,182,550 (Fujitsu).
In addition to the restriction concerning the total number of point codes which can be used within one network, there is a restriction concerning the number of trunk circuits which can belong to any one trunk route. This is because the SS No.7 signalling message field which is used to identify the trunk circuit within the trunk route is of a fixed size. When SS No.7 is used, each national trunk exchange is allocated one and only one point code. This means that there can be one and only one trunk route between any two neighbouring exchanges. The combination of these restrictions means that SS No.7 dictates the maximum number of trunk circuits which can be used between any two neighbouring exchanges.
A conventional architecture as shown in FIG. 1 uses SS No.7 for inter-exchange communication. Fairly complex call processing procedures are used as shown in schematic form in FIG. 3.
An exchange is defined as having a call switching function and call processing functions. The call processing functions are often split, so that they are associated with the incoming half and the outgoing half of the call. The exchange supports signalling systems such as SS No.7 at its boundaries for inter-exchange communication.
In FIG. 3, the local exchange having point code “A” sets up a call to the local exchange having point code “B”, the call is routed via trunk exchanges “X” and “Z”. An alternative route via trunk exchanges “Y” and “Z” is shown. Alternative routes are normally provided as a back-up in case the first choice route is congested or not available. However, two alternative routes with equal priority are quite often provided and the load is shared between them; this is for better resilience. The signalling paths are not shown in FIG. 3, for reasons of clarity.
In FIG. 3, the call processing functions are used to set up and supervise a call from “A” to “B” via exchanges “X” and “Z”, or via exchanges “Y” and “Z”. Although common channel signalling systems such as SS No.7 enable faster call set-up than earlier signalling systems, it is still fairly slow. SS No.7 requires considerable processing resources in each exchange, thus requiring the provision and maintenance of much hardware.
Where expansion of the overall capacity is required, more transit and tandem exchanges might be necessary. However, this increases the amount of processing required per call, which adds to the expense incurred.
SUMMARY OF THE INVENTION
The invention seeks to improve on known systems and methods.
According to a first aspect of the invention, there is provided an exchange comprising:
a plurality of distributed nodes for processing calls from or to a first neighbouring exchange,
two or more of said nodes being arranged to be connected to the first neighbouring exchange by respective trunk routes, and
a point code being assigned to the node-end of each trunk route, wherein the point codes, which are assigned to the node-ends of the separate trunk routes to the first neighbouring exchange, differ from each other.
The exchange of the first aspect of the invention is one single exchange, albeit distributed and consisting of many nodes. From the perspective of neighbouring exchanges, it behaves like one exchange. It is also one exchange in signalling terms because a call can be passed between any pair of its nodes without using any type of inter-exchange signalling.
In the exchange of the invention, signalling point codes are allocated to the node-ends of the trunk routes. Point codes are not allocated to the nodes themselves; and point codes are not allocated to the exchange itself (as would be the situation for conventional exchanges). One node can support more than one signalling point code. The same point code can be allocated to more than one trunk route node-end (providing the trunk routes lead to different neighbouring exchanges).
If there are two or more trunk routes between the first neighbouring exchange and the exchange of the invention, then the first neighbouring exchange sees the exchange of the invention as being two or more exchanges. The first neighbouring exchange identifies each of these trunk routes via its own point code and the point code allocated to the node-end of the trunk route. Therefore the first neighbouring exchange views the other end of the said trunk route as being connected to an exchange which is identified by the point code allocated to the node-end of that trunk route.
The first neighbouring exchange has a choice of trunk routes to the exchange of the invention and, if desired, can allocate them equal priority and use load sharing.
Thus the exchange of the invention can be introduced into a network with a minimum of disruption. The neighbouring exchanges do not need any adaptation in order to be connected to the exchange of the invention.
The exchange of the invention can replace one or more existing exchanges. If it replaces two existing exchanges, and if the first neighbouring exchange was connected to each of the two replaced exchanges via a trunk route, then the first neighbouring exchange will be connected via two trunk routes to the exchange of the invention and will view the exchange of the invention as being two exchanges. No design, implementation or equipment changes will be needed in the first neighbouring exchange. All that will be required is that the first neighbouring exchange updates its data in order to alter the identities of the said two trunk routes.
The exchange of the invention can replace one existing exchange or many existing exchanges. The nodes in the exchange of the invention co-operate with each other and act as one exchange. Therefore when the exchange of the invention replaces several existing exchanges, the total amount of call processing and signalling activity is reduced.
Irrespective of how many existing exchanges the exchange of the invention replaces, the exchange of the invention can use as few as two point codes. The exchange of the invention can be expanded to any size (e.g. by adding more nodes), and still use as few as two point codes. The number of point codes required by the exchange of the invention does not depend upon the total number of nodes in the exchange. The number of point codes required by the exchange of the invention depends on the maximum number of trunk routes between the exchange of the invention and any one neighbouring exchange.
With normal exchanges, there can be one and only one trunk route between any two exchanges and thus there is a maximum number of trunk circuits which can exist between the two exchanges. If desired, there can be several trunk routes between the exchange of the invention and any neighbouring exchange; therefore there is no real restriction on the number of trunk circuits which can exist between the exchange of the invention and any one neighbouring exchange.
Because the exchange of the invention is expandable it can replace many existing exchanges. If this occurs, then because the exchange of the invention can use only a few point codes (normally only two), the number of point codes required by the network can actually be reduced in spite of a possible overall increase in traffic carrying capacity.
The distributed nature of the exchange of the invention means that its nodes are independently operable, and each node is able to continue working in spite of the presence of faults in other parts of the exchange. The nodes can be geographically distributed; this improves the resilience of the exchange to power supply failures and local problems affecting the building in which a node resides. More than one trunk route can be connected from a node to the same neighbouring exchange, providing the node-end of each trunk route is assigned a different point code. This is useful if capacity requirements warrant the use of more than one trunk route.
According to another aspect of the invention, there is provided a switching system for a communication network, the network also comprising a plurality of other devices inter-connected by routes, traffic being routed according to point codes assigned to points in the network, the switching system being connected to a first of said other devices by at least a first and a second of said routes, wherein the switching system comprises:
means for processing traffic on these two or more routes, wherein point codes are assigned to the switching system ends of each of these two or more routes, these point codes being mutually different.
According to another aspect of the invention, there is provided a method of routing traffic across a communication network, the network comprising a plurality of devices inter-connected by routes, point codes being assigned to identify an end of at least some of the routes in the network, wherein at least some of the point codes are re-used in such a manner that the routes are still uniquely identifiable by reference to the two or more point codes assigned to the ends of a respective route, wherein the method comprises the step of:
routing traffic across the network on the basis of the point codes.
According to another aspect of the invention, there is provided a method of routing traffic across a communication network, the network comprising a plurality of devices inter-connected by routes, codes being assigned to identify an end of at least some of the routes in the network, the method comprising the step of:
routing traffic to a subset of the devices in the network, wherein the ends of routes arriving at the subset of devices are identified by a set of two or more of said codes, none of which are re-used outside the subset, wherein at least some of the codes in the set are re-used in the subset, and wherein the step of routing traffic to the subset is carried out by routing the traffic to any of the set of codes.
Preferred features are set out in dependent claims. Preferred features may be combined as would be appreciated by a skilled person.