Method and apparatus for learning network behavior trends and predicting future behavior of communications networks

• US 5,987,442 A
• Filed: 10/24/1997
• Issued: 11/16/1999
• Est. Priority Date: 02/02/1995
• Status: Expired due to Term

• Current Assignee: Computer Associates Think Inc. (Broadcom, Inc.)
• Original Assignee: Cabeltron Systems, Inc.
• Inventors: Lewis, Lundy, Datta, Utpal
• Primary Examiner(s): Lintz, Paul R.
• Assistant Examiner(s): Shah, Sanjiv
• Application Number:
  US08/960,076
• Time in Patent Office: 753 Days
• Field of Search: 706/15, 706/16, 706/25, 706/10, 706/20, 364/488, 395/500
• US Class Current: 706/10
• CPC Class Codes: H04L 43/00 Arrangements for monitoring or testing data switching networks
• CPC Class Codes: H04L 43/045 for graphical visualisation of monitoring data
• CPC Class Codes: H04L 43/0817 by checking functioning
• CPC Class Codes: H04L 43/16 Threshold monitoring

Abstract

Apparatus and method for learning current network behavior and predicting future behavior which utilizes a state transition graph. The graph includes nodes which represent network states, and arcs which represent trends in observable network parameters that result in a transition from a current state to another state. For example, a watch service may be instituted on multiple ports of a router, and the observed network traffic on the ports over time may be transformed into a state transition graph that represents network behavior. The network states may be labeled such as "good", or "bad", etc., according to a predetermined performance criteria. Once a state transition graph is constructed, the system may then monitor the current state and current trends of the network parameters in order to predict and display future network states. The system may include an automatic warning signal for alerting a user that the network is headed in the direction of a problematic state. The prediction of future network behavior may be made from the state transition graph, the current state of the network, and the current network trends.

Claims

1. A method of determining behavior patterns of a communications network from accumulated network parameters, the method including the steps of:

a) reading accumulated network parameters representing data traffic over a multiple segment communications network with the segments connected by network couplers;

b) determining a set of candidate states of the communications network based upon the accumulated network parameters, each said state being delineated by a learning algorithm;

c) determining a set of candidate transitions among the set of candidate states using the network parameters; and

d) displaying to a user, one or more of the set of candidate states and the set of candidate transitions corresponding to behavior patterns of the multiple segment communications network.

2. The method of claim 1, further including:

measuring current network parameters for a predetermined time period;

determining a current state of the network from the current network parameters and the set of candidate states; and

displaying to a user the current state of the communications network.

3. The method of claim 2, further including:

determining current trends of the network from the current network parameters and the set of candidate transitions;

predicting a set of candidate next states that the network may enter from the set of candidate states, the set of candidate transitions, the current state, and the current trends; and

displaying to a user at least one of the current trends and the set of candidate next states.

4. The method of claim 3, further including:

labeling at least one of the set of candidate states, the set of candidate transitions, the current state, the set of candidate next states and the current trends according to a performance criterion.

5. The method of claim 3, further including:

designating at least one of the set of candidate states as a problematic state; and

warning a user when any of the set of candidate next states is a problematic state.

16. The method of claim 1, wherein the step of determining the set of candidate transitions includes determining the set of candidate transitions based upon the accumulated network parameters.

6. A method of determining behavior patterns of a communications network, the method including the steps of:

a) measuring accumulated network parameters representing data traffic, over a multiple segment communications network with the segments connected by network couplers, for a first predetermined time period;

b) storing the accumulated network parameters in a data repository;

c) reading the accumulated network parameters from the data repository;

d) determining a set of candidate states of the network based upon the accumulated network parameters, each said state being delineated by a learning algorithm;

e) determining a set of candidate transitions among the set of candidate states using the network parameters; and

f) displaying to a user, one or more of the set of candidate states and the set of candidate transitions corresponding to behavior patterns of the multiple segment communications network.

7. The method of claim 6, further including:

measuring current network parameters for a second predetermined time period;

determining a current state of the network from the current network parameters and the set of candidate states; and

displaying to a user the current state of the communications network.

8. The method of claim 7, further including:

determining current trends of the network from the current network parameters and the set of candidate transitions;

predicting a set of candidate next states that the network may enter from the set of candidate states, the set of candidate transitions, the current state, and the current trends; and

displaying to a user at least one of the current trends and the set of candidate next states.

9. The method of claim 8, further including:

labeling at least one of the set of candidate states, the set of candidate transitions, the current state, the set of candidate next states, and the current trends according to a performance criterion.

10. The method of claim 9, further including:

designating at least one of the set of candidate states as a problematic state; and

warning a user when any of the set of candidate next states is a problematic state.

17. The method of claim 6, wherein the step of determining the set of candidate transitions includes determining the set of candidate transitions based upon the accumulated network parameters.

11. An apparatus that reads accumulated network parameters from a data repository of a network management system, and that provides a description of behavior patterns of a communications network, the apparatus comprising:

a data reader, coupled to the data repository, that reads in accumulated network parameters representing data traffic over a multiple segment communications network with the segments connected by network couplers;

a behavior analyzer, coupled to the data reader, that determines a set of candidate states of the communications network based upon the accumulated network parameters and determines a set of candidate transitions between the set of candidate states using the network parameters, each said state being delineated by a learning algorithm; and

a states display, coupled to the behavior analyzer, that communicates to a user the set of candidate states and the set of candidate transitions corresponding to behavior patterns of the multiple segment communications network.

12. The apparatus according to claim 11, further comprising:

a monitor, coupled to the network, that measures current network parameters for a predetermined time period;

a status analyzer, coupled to the monitor and the behavior analyzer, that determines a current state of the network from the current network parameters and the set of candidate states; and

a current status display, coupled to the status analyzer, that communicates to a user the current state of the communications network.

13. The apparatus according to claim 12, further comprising:

a trends analyzer, coupled to the monitor and the behavior analyzer, that determines current trends of the network from the current network parameters and the set of candidate transitions;

a behavior predictor, coupled to the status analyzer, the trends analyzer, and the behavior analyzer, that determines a set of candidate next states that the network may enter from the set of candidate states, the set of candidate transitions, the current state, and the current trends; and

a predictor display, coupled to the trends analyzer and the behavior predictor, that communicates to a user at least one of the current trends and the set of candidate next states.

14. The apparatus according to claim 13, further comprising:

a state identifier, coupled to at least one of the behavior analyzer, the status analyzer, and the behavior predictor, that labels at least one of the set of candidate states, the set of candidate transitions, the current state, the set of candidate next states and the current trends according to a performance criterion.

15. The apparatus of claim 13, further comprising:

a state identifier that designates at least one of the set of candidate states as a problematic state; and

a warning unit, coupled to the state identifier, that warns a user when any of the set of candidate next states is a problematic state.

18. The apparatus of claim 11, wherein the behavior analyzer is constructed and arranged to determine the set of candidate transitions based upon the accumulated network parameters.

19. An apparatus for determining behavior patterns of a communications network from accumulated network parameters, comprising:

means for reading accumulated network parameters representing data traffic over a multiple segment communications network with the segments connected by network couplers;

means for determining a set of candidate states of the communications network based upon the accumulated network parameters, each said state being delineated by a learning algorithm;

means for determining a set of candidate transitions among the set of candidate states using the network parameters; and

means for displaying to a user, one or more of the set of candidate states and the set of candidate transitions corresponding to behavior patterns of the multiple segment communications network.

20. The apparatus of claim 19, further comprising:

means for measuring current network parameters for a predetermined time period;

means for determining a current state of the network from the current network parameters and the set of candidate states; and

means for displaying to a user the current state of the communications network.

21. The apparatus of claim 20, further comprising:

means for determining current trends of the network from the current network parameters and the set of candidate transitions;

means for predicting a set of candidate next states that the network may enter from the set of candidate states, the set of candidate transitions, the current state, and the current trends; and

means for displaying to a user at least one of the current trends and the set of candidate next states.

22. The apparatus of claim 21, further comprising means for labeling at least one of the set of candidate states, the set of candidate transitions, the current state, the set of candidate next states and the current trends according to a performance criterion.

23. The apparatus of claim 21, further comprising:

designating at least one of the set of candidate states as a problematic state; and

warning a user when any of the set of candidate next states is a problematic state.

24. The apparatus of claim 19, wherein the means for determining the set of candidate transitions includes means for determining the set of candidate transitions based upon the accumulated network parameters.

25. An apparatus for determining behavior patterns of a communications network, the apparatus comprising:

means for measuring accumulated network parameters representing data traffic, over a multiple segment communications network with the segments connected by network couplers, for a first predetermined time period;

means for storing the accumulated network parameters in a data repository;

means for reading the accumulated network parameters from the data repository;

means for determining a set of candidate states of the network based upon the accumulated network parameters, each said state being delineated by a learning algorithm;

means for determining a set of candidate transitions among the set of candidate states using the network parameters; and

means for displaying to a user, one or more of the set of candidate states and the set of candidate transitions corresponding to behavior patterns of the multiple segment communications network.

26. The apparatus of claim 25, further comprising:

means for measuring current network parameters for a second predetermined time period;

means for determining a current state of the network from the current network parameters and the set of candidate states; and

means for displaying to a user the current state of the communications network.

27. The apparatus of claim 26, further comprising:

means for determining current trends of the network from the current network parameters and the set of candidate transitions;

means for predicting a set of candidate next states that the network may enter from the set of candidate states, the set of candidate transitions, the current state, and the current trends; and

means for displaying to a user at least one of the current trends and the set of candidate next states.

28. The apparatus of claim 27, further comprising means for labeling at least one of the set of candidate states, the set of candidate transitions, the current state, the set of candidate next states, and the current trends according to a performance criterion.

29. The apparatus of claim 28, further comprising:

means for designating at least one of the set of candidate states as a problematic state; and

means for warning a user when any of the set of candidate next states is a problematic state.

30. The apparatus of claim 25, wherein the means for determining the set of candidate transitions includes means for determining the set of candidate transitions based upon the accumulated network parameters.