Computer platform alarm and control system

• US 6,446,058 B1
• Filed: 04/26/1999
• Issued: 09/03/2002
• Est. Priority Date: 04/26/1999
• Status: Expired due to Term

• Current Assignee: AT&T Corporation (AT&T, Inc.)
• Original Assignee: AT&T Corporation (AT&T, Inc.)
• Inventors: Brown, Abby H.
• Primary Examiner(s): STARKS, WILBERT L
• Application Number:
  US09/299,018
• Time in Patent Office: 1,226 Days
• Field of Search: 706/60, 379/9, 379/10, 379/29
• US Class Current: 706/60
• CPC Class Codes: G06N 5/045 Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence
• CPC Class Codes: H04L 41/16 using machine learning or artificial intelligence
• CPC Class Codes: H04M 3/10 Providing fault- or trouble-signals
• CPC Class Codes: H04M 3/2263 Network management

Abstract

A computer network alarm and control system for use in a network comprising a plurality of platforms susceptible to failure comprises a notification system coupled to the network of platforms for receiving an alarm signal from the platforms and outputting a notification identifying a platform indicating an alarm, a qualification of the alarm and a severity of the alarm and an expert system including an inference engine coupled to the notification system, the inference engine having an associated knowledge base for outputting an appropriate resolution to the notification system. An optional explanation facility system communicates with the inference engine of the expert system, the explanation facility system, responsive to a query, outputting an explanation for the appropriate resolution. The expert system is preferably frame-based, having frames for software, hardware and middleware faults and errors, each frame having slot names and values, the slot names including the identity of the platform (or node), symptom, severity, resolution and the like.

Claims

1. A computer network alarm and control system for use in a network comprising a plurality of platforms susceptible to failure comprising:

a notification system coupled to said network of platforms for receiving an alarm signal from said platforms and outputting a notification identifying a platform indicating an alarm, a qualification of said alarm and a severity of said alarm; and

an expert system including an inference engine coupled to said notification system, said inference engine having an associated knowledge base for outputting an appropriate resolution to said notification system.

2. A computer network alarm and control system as recited in claim 1 further comprising:

an explanation facility system for communication with said inference engine, the explanation facility system, responsive to a query, outputting an explanation for said appropriate resolution.

3. A computer network alarm and control system as recited in claim 1 wherein said notification system is coupled to said plurality of platforms by links, said notification system for periodically receiving fault and error data from said platforms in real time over said links.

4. A computer network alarm and control system as recited in claim 1 wherein said inference engine is frame-based, each frame comprising at least one slot name and slot value, frames being defined for each of hardware, middleware and software.

5. A computer network alarm and control system as recited in claim 4 wherein slot names include nodal identification, symptom, severity and resolution.

6. A computer network alarm and control system as recited in claim 4 wherein said at least one slot name comprises at least four slot names representing nodal identification, symptom, severity and resolution.

15. The computer network alarm and control system of claim 4, wherein the frames include hierarchies of classes, objects and attributes of objects.

7. A method for use in a computer network comprising a plurality of platforms susceptible to failure, the method for outputting a resolution responsive to receipt of an alarm signal at a notification system and outputting a resolution from an inferencing engine comprising the steps of:

receiving an alarm signal from a platform at said notification system;

identifying the platform;

diagnosing the cause of the alarm signal;

matching data representing said cause to a plurality of frames of a knowledge base at said inferencing engine; and

outputting a resolution responsive to a best match of said cause data to a selected one of said frames of said knowledge base.

8. A method as recited in claim 7 further comprising the step of:

receiving a query for an explanation of said resolution; and

outputting said explanation for presentation via a presentation terminal.

9. A method as recited in claim 7 wherein said frames each comprise a plurality of slots for invoking system calls to one of rules or procedural attachments to a corresponding frame.

10. A method as recited in claim 9 wherein the frames are defined in Prolog.

11. A method as recited in claim 7 wherein said inference engine is frame-based, each frame comprising at least one slot name and slot value, frames being defined for each of hardware, middleware and software.

12. A method as recited in claim 7 further comprising a method of determining outage time comprising the steps of:

validating an outage, obtaining a time stamp when a system logs the outage and obtaining a time stamp for a successful completion of a resolution; and

calculating an outage time from said time stamps.

13. A method as recited in claim 7 wherein said resolution output is responsive to a severity level input from said notification system.

14. A method as recited in claim 13 wherein said resolution is further responsive to identification of a platform and a reported symptom.

16. The method of claim 7, wherein said plurality of frames include hierarchies of classes, objects and attributes of objects.