Data processing: artificial intelligence – Knowledge processing system – Knowledge representation and reasoning technique
Reexamination Certificate
1998-03-18
2001-06-05
Powell, Mark R. (Department: 2762)
Data processing: artificial intelligence
Knowledge processing system
Knowledge representation and reasoning technique
C706S010000, C706S021000, C706S045000
Reexamination Certificate
active
06243697
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to telecommunications networks, and in particular to the detection of undesirable interactions between different services running on a network.
RELATED ART
Telecommunications networks are increasingly required to offer customers services in addition to basic call-handling. The development of novel network architectures, such as the IN (intelligent network) architecture, together with developments in computing platforms for telecommunications systems, make it potentially possible to offer customers a large portfolio of additional services to select from. However, as the number of services increases, and as services become available from independent service providers in addition to the network operator, then service feature interaction becomes a serious problem. It is often found that features offered by one service interact in an unwanted manner with features of other services. For example, a voice messaging service, such as BT's CallMinder service, may have as one of its standard features a behaviour such that incoming calls are diverted to the messaging service whenever the called line is busy. Another available service, Call Waiting, handles the same condition, namely the called number being busy, in an entirely different fashion. The Call Waiting service transmits an alert tone to the user and gives the user the option of interrupting the on-going call to speak to the new caller. It can be seen that if a customer wanted to subscribe to both services, then there is conflict between the service features which needs to be resolved. Otherwise, it would be necessary to bar the provisioning of both of these services to a customer, with a consequent loss in utility to the customer, and loss of revenue to the service provider.
Conventionally, during the planning of new services for a telecommunications network, attempts have been made to detect in advance any interaction problems by writing English-language specifications of the service features. Using these specifications a paper “walk through” of the services is then conducted, with the design engineer going step by step through the different services and spotting any interaction problems. This is a time-consuming procedure which can never be completely reliable, leaving the possibility that unforeseen interactions will occur when the service is deployed.
Some attempts have been made previously to automate the detection of interaction during the design phase. For example, International patent application WO95/22231 discloses a method of detecting service interactions which uses formal specifications of the additional services. The algorithm uses information that is specific to the services being tested which needs to be rewritten every time a change is made to one of the service features. Moreover, the approach adopted requires that a formal model should be prepared for every service feature which is handled. The preparation of such formal models is a difficult and time-consuming task requiring a high level of expertise.
SUMMARY OF THE INVENTION
According to a first aspect of the present invention, there is provided a system for detecting interaction between services running on a telecommunications network comprising:
a computer expert system including:
a) a data store programmed with data representing attributes of service features;
b) a rule store programmed with rules which relate feature attributes to interaction behaviours; and
c) an inference engine which is connected to the data store and to the rule store and which is arranged to process the data and the rules, thereby detecting any interaction between the services.
The present invention adopts a radically new approach to the detection of service feature interaction, through the use of an expert system. The traditional domain of use of expert systems is for diagnostic classification problems involving data which is essentially static. A well known example is the identification of a particular bacterium from a series of statements about its properties and appearance. In this domain, expert system technology has a proven track record in reproducing and sometimes surpassing human expert performance for the same problem. It is not however been thought possible hitherto to apply such techniques to the problem domain of the present invention. Feature interaction in a telecommunications network is an essentially time-related phenomen and so on the face of it not suitable for expert system techniques. The present inventors have found however that with an appropriate knowledge representation, expert systems can successfully be used for feature interaction detection. This provides a dramatic reduction in the time required to uncover service interworking problems, coupled with increased flexibility. The separation between the knowledge representation and the inference rules, means that changes or additions to the services can readily be assessed simply by making corresponding changes to the knowledge representation in the data store. In this way, the system is quickly able to detect any problems arising from the new features. By contrast with the prior art systems there is no need to repeat an entire “walk-through” from scratch.
The system of the present invention may run, for example, at a local switch in a telecommunications network to aid the detection and management of interaction problems as they occur. Alternatively, or in addition, the system may be used during the development of a new service to detect any interaction problems prior to the deployment of the service.
In the case of a system used at run-time in the network, then the output of the system may be fed to a control system for modifying the behaviour of the network in order to remove or ameliorate the detected interaction problem. The control system may, for example, modify the stored profile for a customer in order to disable one or more service features. A more sophisticated control system might initiate a dialogue with the customer to allow the customer to determine a default behaviour for the network. For example, in the case of the Call Minder or call waiting services, the user might be given the option of selecting the Call Waiting response, that is the transmission of an alert tone, rather than the Call Minder response, that is the transfer of the incoming call to a messaging service.
Preferably the expert system data store includes a plurality of objects including, for each service feature which is represented in the data store, a set of objects corresponding to different respective state transitions of the feature.
The term “object” is used in this document in the sense of object oriented design/programming (OOD/OOP) methodologies.
As discussed in further detail below, the choice of an appropriate structure for organising the data in the data store is critical in maximising the efficiency of the interaction detection system. The inventors have found that the combination of the use of an object-based structure and a state transition representation of the service feature offers significant advantages both in efficiency of operation and in ease of development and modification of the detection system. The use of sets of objects representing different state transitions makes it possible to capture the characteristics of a service feature in a form which is well-adapted for processing by the inference engine.
Preferably each of the said set of state transition objects includes a sequence number corresponding to the position of the respective state transition in the sequence of execution of the feature and at least some of the rules in the rule store reason over the values of the sequence numbers. Preferably the objects are arranged in a hierarchy of superclasses and subclasses of the superclasses, and some of the rules reason over superclasses and others of the rules reason over subclasses.
The use of objects belonging to a hierarchy of classes, combined with rules which operate at different levels of the class hierarchy further increases the flexibi
British Telecommunications public limited company
Nixon & Vanderhye P.C.
Powell Mark R.
Starks Wilbert
LandOfFree
Detecting service interactions in a telecommunications network does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Detecting service interactions in a telecommunications network, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Detecting service interactions in a telecommunications network will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2500818