Endre Skolt
This paper gives an introduction to the ongoing standardisation activity on Intelligent Networks (IN). It contains an overview of the first available standard: Capability Set 1 (CS-1). The standard defines the service requirements for CS-1 and introduces the concept of service independent building blocks (SIBs). It also covers call modelling and functional architectural aspects.
The paper also gives an overview of preliminary plans for the next standardisation phase of IN, Capability Set 2.
Standards in the Intelligent Network area are motivated by the interests of telecommunication service providers to meet existing and potential market needs for services in a rapid and cost effective manner. Another important motivating factor for IN standardisation is that the service providers seek to improve the quality of their service offer and to reduce the cost of network operations and management.
One of the objectives of Intelligent Networks (IN) is to allow the inclusion of additional capabilities to facilitate service and network implementation independent provisioning of services in a multi-vendor environment. That means to allow service providers to define their own services independently of equipment suppliers, and in the same way allow network operators to install switches and databases supplied by different manufacturers.
The target IN will be applicable to a wide variety of networks including Public Switched Telephone Networks (PSTN), Public Land mobile Networks (PLMN), Packet Switched Public Data Networks (PSPDN) and Integrated Services Digital Network (ISDN), both Narrowband ISDN (N-ISDN) and Broadband ISDN (B-ISDN).
CS-1 is the first stage of standardisation of the Intelligent Network as an architectural concept for the creation and provision of telecommunication services. It is, however, important to stress that CS-1 only contains a subset of IN capabilities. CS-1 can be viewed as follows:
Traditionally, standardisation has been an activity which implied solutions based on existing technology. In the case of IN the standardisation has become closer to a research project where the results are advancing the current technology. However, due to the project size and the large number of participating organisations, compromises are necessary in order to make progress, and the standards may therefore technically not represent the optimal solutions for each service provider.
CCITT SG XI started work on IN in 1989. There were different opinions concerning the time frame for which the group wanted to achieve implementable recommendations. Two opinions were confronted: On the one hand, strong market needs require implementable specifications in a very short time frame. It was also stressed that the market does not have time to wait several years for pilot trials.
On the other hand, it was argued that in order to achieve viable solutions it would be necessary to work on a long term basis.
As the work progressed it became clear that most of the participants favoured the first approach and aim for specifications already in 1992. (CS-1 will formally be approved in March 1993.) In Figure 1 some milestones of the work on IN is shown.
CCITT is a consultative committee which produces recommendations for the International Telecommunication Union (ITU). CCITT is organised into study groups, and they are open for all parties in the telecommunication society. A number of equipment manufacturers, service providers, network operators and organisations representing the customers from all over the world have participated in the specification of CS-1.
The large number of participants imply that CS-1 is a compromise of many interests. There is, however, no doubt that Bellcore's early IN studies have had considerable influence on CS-1. In addition, the equipment manufacturers have been very active in order to prevent the standards to impose too many changes to current switching technology.
As mentioned above, CS-1 contains only a subset of the target IN architecture, but will give the telecommunication players a common platform for future evolution. In the first stage of the work much effort has been focused on the modelling of the IN concept and a conceptual model for IN has been developed to structure the concept. The model consists of four planes where each plane represents a theoretical view of the services and network functions as follows:
CS-1 provides an IN platform, which is the base for provision of services. This platform has been developed on the basis of a set of benchmark services and service features. The services and service features of CS-1 are based on a set of Service Independent building Blocks (SIBs). The SIBs are the service components required for creation and composition of services.
An important accomplishment of CS-1 is the definition of communications protocols for the interface between the Service Switching Function (SSF) and the Service Control Function (SCF) (see definitions on page 49). CS-1 also contains other protocol standards. These will be discussed later in this paper.
The main results of CS-1 is in the area of service execution, that means the rules and procedures to establish, maintain and release a call. Service creation and service administration have not been dealt with in the same depth but may be a major working area in the next stage of IN standardisation (CS-2).
In this section I have summarised some benefits the different telecommunication parties (see Figure 2) will gain from CS-1:
The separation of switching functions, service logic, specialised resources and subscriber data enables the network operator to purchase equipment from different suppliers. Such equipment includes Service Control Points (SCPs), Service Switching Points (SSPs), Intelligent Peripherals (IPs) and Service Data Points (SDPs). In addition CS-1 will provide efficient network routing and give savings in network and transmission resources. However, the savings must be paid for by an increase in signalling traffic.
Flexible provision of services and the capability to offer services like Universal Personal Telecommunication (UPT) and Virtual Private Networks (VPN) across several neworks operated by different service providers will also be possible. However, initially there will be several restrictions on utilising the full capabilities of these services.
One fundamental element of the IN concept is the service independent building blocks (SIBs). By combining SIBs the service providers and network operators may offer a large number of services. However, CS-1 only offers some guidelines for the introduction and deployment of services into the network.
End users and service subscribers will through advanced user-network interaction have access to a large variety of services including customised service profiles, flexible charging and user control.
The development cost in software and hardware for public network systems are considerable. Therefore, it is vital for the equipment manufacturers to agree on world wide standards in order to have large markets for their products. With CS-1 they have accomplished a global standard.
The service plane of the IN conceptual model illustrates the service from a user's perspective. A service may be composed of several service features or service components. A list of service features for CS-1 is shown in Table 1. See Q.1211 for a more comprehensive description.
The type of services which can be offered in CS-1 is classified as type A services. Services that require IN handling but cannot be supported by CS-1 are called type B. In CCITT recommendation Q.1211 the following definition is given:
"The CS-1 capabilities are intended to support service and service features that fall into the category of single ended, single point of control services referred to as type A, with all other services placed in a category called type B."
Single-ended service features mean that these features only apply to one party in a call and are independent of any other parties that may be participating in the call. Single point of control means that a control relationship in a call can only exist to one Service Control Function at any point in time.
A conference call as a type A service will provide the participants very limited flexibility in the control and execution of the call. For example, the A-party will have to keep the control during the complete lifetime of the call and the associations of the parties physically take place in the switches controlled by only one Service Control Function.
For conference calls as type B the associations of the parties may physically take place in the switches controlled by more than one Service Control Function. It may also be offered that parties are added to or dropped off and that the control of the call can be transferred from one party to another. A conference call example where the involved parties are connected to switches in different networks and where service control interaction is necessary is shown in Figure 3.
In complex services the SCF will need rules to handle feature interaction2) between parties in a call. Since this aspect is not covered by CS-1, service providers and network operators need to rely on network and supplier specific solutions. The feature interaction problem is illustrated by an example where a Freephone service interacts with a Call Completion on Busy Subscriber (CCBS) service (see Figure 4). The calling user dials the Freephone number and a busy tone is returned. Then the calling user activates the CDBS service. In the network the following procedure will take place:
When the network receives the Freephone number, the called party's Freephone service logic program3) will be invoked. The service logic program will convert the Freephone number to a network routing number and the network will use this number to route the calls to the Freephone destination. When the Freephone user is busy the terminating exchange will return a busy signal. When receiving the busy signal from the terminating exchange the calling user activates CCBS causing the network to start monitoring the called party's access. When the called party's access becomes free, the exchange will alert the originating exchange and the call will automatically be set up towards the called party.
In this case the two services will be independently executed since the Freephone service logic program belongs to the called party and the CCBS service logic program belongs to the calling user. The interaction problem occurs because the terminating exchange will view the second attempt as an ordinary call (not a Freephone call).
There is a number of reasons why type B services are not included in CS-1. Type B services will require a complexity which has not been possible to explore within the time frame of CS-1. Type B services may also involve manipulation of abstract switching connections like legs4) and connection points. Existing switching equipment is not designed for such applications.
The switching manufacturers have been reluctant to introduce new software modelling in order to support these advanced capabilities.
On the other hand, type A services can be implemented by relatively simple control relationships between the switch and service logic database. In contrast the type B services may require sharing of connection control and solutions based on distributed processing with large protocol complexity.
The global functional plane of the IN conceptual model is the service provider's view of the telecommunication network. This plane describes the service creation process by use of SIBs. For CS-1 fourteen SIBs have been defined, but rules and methodology to combine the SIBs have not been defined. A platform for creation and execution of services based on SIBs may be developed for CS-2. A list of all the SIBs defined for CS-1 is provided in Table 2.
SIBs are abstract representations of network capabilities that will exist in an IN structured network. As the name of these building blocks indicates, a SIB is independent of user services and the technology the services are developed from.
CS-1 contains stage 1 description, i.e. description of what each SIB does, and stage 2 description, i.e. a description of which functional entities the SIB involves and information flows between these functional entities. A stage 1 description of the Translate SIB is given in Figure 5.
The distributed functional plane contains the IN functional architecture and provides a network operator's view of the network. The functional architecture consists of functional entities (FE) where each FE represents a grouping of network functions, for example service control logic, specialised resources, service data, connection control, etc. The distributed functional plane also defines relationships between functional entities.
The following functional entities required for call handling are defined:
The functional architecture is shown in Figure 6.
The call control aspects involve the functional entities CCF, SSF and SCF. Some key principles for CS-1 call control are:
In a large number of services there will be a need for user-network interaction in order to provide a user-friendly access. This interaction will be suported by a variety of text messages (ISDN) and voice announcements. For this purpose the specialised resource function or an intelligent peripheral (physical network node) is used. An important principle in user-network interaction is that the Service Control Function will have the capability to suspend and resume service processing on behalf of the calling and called party.
With the enhanced capabilities for user-network interaction, user control will be offered to the customer in order to manage the services. Customer control includes the functionality to interrogate, modify and delete records in the customer's service profile. The customer control procedure primarily addresses the functional entities SSF, SCF, SDF and SRF. A customer control procedure will normally take place outside the context of a call.
The CS-1 standards do not cover service management aspects, but the entities responsible for management have been identified:
The functional architecture including the service management related functional entities is shown in Figure 7.
For CS-1 thirteen functional relationships or interfaces between functional entities have been identified. The functional relationships are shown in figure 8. In addition four types of control relationship have been defined in order to distinguish the type of physical interfaces. They are defined as follows:
For CS-1 the protocols between the Service Switching Function and the Service Control Function (SSF-SCF), the Service Control Function and the Specialised Resource Function (SCF-SRF) and the Service Control Function and the Service Data Function (SCF-SDF) are standardised. These protocols are all of the service control relationship type. With reference to the OSI layer model, they will be implemented as application layer protocols.
In addition to the work on interface standards for IN, CCITT has defined application layer structure for application protocols which is applicable to the IN protocols. For all the CS-1 specified protocols the same underlying protocols can be used, e.g. Transaction Capabilities Application Part (TCAP) or application protocols supporting remote operations.
CS-1 has provided recommendations to be used within a single network in a multi-vendor environment. Therefore, the network operators will have to rely on bilateral or multilateral agreements in order to offer services across networks, e.g. Freephone services, Universal Personal Telecommunication services and Virtual Private Network services.
However, the interface between Service Control Function and Service Data Function may be implemented as an inter-network interface. An application of this interface is seen for the offering of credit calling services, where the Service Data Function may be a database which contains authentication data, e.g. personal identification number (PIN) codes in combination with account numbers.
The Service Control Function to Service Switching Function relationship is not considered as an inter-network interface. This interface would require additional security capabilities in order to be taken into account. Work on Open Network Provision (ONP) may have impact on whether or not this interface will be considered in future capability sets.
The interface between two Service Control Functions would violate the "single point of control" which is a requirement for the type of services to be offered within the scope of CS-1, thus not standardised. In the future evolution of IN this relationship must be considered, in particular in order to solve the service feature interaction problem.
The interface between two Service Data Functions is not standardised for CS-1. However, this relationship may be important in mobility applications for location handling and optimal distribution of data.
Interfaces between Service Management Functions have not been considered as an inter-network interface.
An illustration of the possible interfaces for network interworking is given in Figure 9.
It is a common view that it will take many years to accomplish the target IN which is characterised by attributes such as efficient use of network resources, subscriber control of service attributes and customisation of services, modularisation and reusability of network functions, integrated service creation and implementation, standardised management of service logic, etc.
The way that has been chosen to reach this goal is to define capability sets which are to be subject of standardisation activity. To ensure a smooth evolution towards the target IN the recommendations produced must fulfill two important criteria:
At the moment there are no fixed plans for when the future capability sets will be ready. However, it has been proposed to issue results from the standardisation activity every second year. That would imply CS-2 recommendations in 1994-95, CS-3 recommendations in 1996-97, etc. (see Figure 10).
CS-1 is the first stop towards the target IN. Even though the technical work has formally been completed there is a need for a maintenance activity of CS-1. This will include removal of inconsistencies in the current version as implementation experience is progressing. This work will be done in parallel with the initiation of the next capability set.
At this stage detailed planning of the next capability set, CS-2, has not yet started, however, it is likely that CS-2 will include studies in a number of areas such as:
Modelling of the IN concept will be an important part of the work and may in addition to the above mentioned areas be influenced by the ongoing activities in broadband and mobile radio.
The following outlines some preliminary thoughts concerning the work items for CS-2:
The future networks may be structured according to Intelligent Networks and Telecommunication Management Networks (TMN), where TMN is the concept for support of the network operator's and service provider's management requirements. This includes procedures for planning, provisioning, installing, operation and administration of telecommunication services. Since the IN concept also incorporates management aspects, some work has to be done in order to align the two concepts. The work has already started, and for CS-2 this will be a prioritised work item. Part of the work will include modelling of service management functional entities illustrated in Figure 7.
One of the most complex areas of the work on IN is service interaction. Service interaction is understood as the mutual influence services may have on each other, and occurs when the execution of a service modifies the behaviour of the execution of another service. In this paper a service interaction example has been illustrated for a Freephone service and Call Completion to Busy Subscriber service, see Figure 4. In order to provide advanced services for CS-2, service and feature interaction must be studied in detail.
The CS-1 recommendation set focuses on network internal aspects, and is very little concerned with network interworking aspects. However, in order to accomplish standardised solutions for international services across several networks some work has to be done. For CS-2 it is proposed to initiate work in this area, both network interworking between public networks and public and private networks.
Security aspects will also have to be studied in CS-2. Basically there are two types which are important for IN:
Access control contains the user identification, user authentication and user authorisation by the network, where the user may be end-users, service providers and network operators. Data control relates to the control of input data, which has been introduced by a user.
1 CCITT Recommendation Q.1204, March 1992
2 CCITT Recommendation Q.1211, March 1992
3 CCITT Recommendation Q.1213, March 1992
4 CCITT Recommendation Q.1214, March 1992
5 CCITT Recommendation Q.1219, March 1992
6 Skolt, E. Intelligente Nett: Kort Innføring. Kjeller, Norwegian Telecom Research, 1991. (U1/91)