Granted Patent B2

US 12,452,141 · App. 18/493,369 · Granted Oct 21, 2025

Inventors: Eric A Voit (Bethesda, MD); John Michael Lake (Cary, NC); Carlos M. Pignataro (Cary, NC)

Assignee: Cisco Technology, Inc.

H04L41/147H04L41/16

Abstract

Claims (52)

1. A method performed at least in part by a network device for seasonal resource provisioning for an ultra-short-term forecast, the method comprising:

receiving, from a network controller, a selection of a time series algorithm, for determining capacity right-sizing of a local resource, from a set of time series algorithms, wherein the time series algorithm is a computationally efficient algorithm based at least in part on current local traffic conditions and wherein a right-size capacity is at least an amount of the local resource needed to ensure proper network functioning;

receiving, from the network controller, an instruction to run the time series algorithm when a local traffic condition trigger event occurs, wherein the instruction includes a determination of parameter values associated with the time series algorithm selection to be used in the time series algorithm;

receiving, from the network controller, a determination of a number of data points for input to the time series algorithm, wherein data points are a sequence of values representing an amount of the local resource used by the network device at a point in time that are collected at predetermined time intervals;

determining, based at least in part on a calculation of the time series algorithm using the number of data points and parameter values and in response to the local traffic condition trigger event occurring, the right-size capacity of the local resource for the network device; and

providing the right-size capacity of the local resource for the network device.

2. The method of claim 1 , further comprising receiving, from the network controller, an acceptable failure rate parameter for the time series algorithm, the acceptable failure rate parameter based at least in part on current local traffic conditions.

3. The method of claim 1 , further comprising:

determining whether a current data point value collected is greater than a previous data point value;

based at least in part on determining that the current data point value is not greater than the previous data point value, continue providing the right-size capacity of the local resource for the network device; and

based at least in part of determining that the current data point value is greater than the previous data point value, execute a new time series calculation to determine a new right-size capacity of the local resource for the network device.

4. The method of claim 1 , wherein an individual time series algorithm of the set of time series algorithms includes pre-provisioned training data.

5. The method of claim 1 , further comprising uploading, by the network device and to the network controller, trained data for distribution to other network devices.

6. The method of claim 1 , further comprising:

storing local time series conditions data in the network device that is collected prior to an unexpected exceeding of capacity of the local resource on the network device; and

providing the local time series conditions data as backpropagation training data for the set of time series algorithms for the network device.

7. The method of claim 1 , further comprising including external indications of near future events for determining the right-size capacity of the local resource for the network device.

8. A system comprising:

one or more processors; and

one or more computer-readable media storing computer-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:

receiving, by a network device and from a network controller, a selection of a time series algorithm, for determining capacity right-sizing of a local resource, from a set of time series algorithms, wherein the time series algorithm is a computationally efficient algorithm based at least in part on current local traffic conditions and wherein a right-size capacity is at least an amount of the local resource needed to ensure proper network functioning;

receiving, by the network device and from the network controller, an instruction to run the time series algorithm when a local traffic condition trigger event occurs, wherein the instruction includes a determination of parameter values associated with the time series algorithm selection to be used in the time series algorithm;

receiving, by the network device and from the network controller, a determination of a number of data points for input to the time series algorithm, wherein data points are a sequence of values representing an amount of the local resource used by the network device at a point in time that are collected at predetermined time intervals;

determining, based at least in part on a calculation of the time series algorithm using the number of data points and parameter values and in response to the local traffic condition trigger event occurring, the right-size capacity of the local resource for the network device; and

providing, the right-size capacity of the local resource for the network device.

9. The system of claim 8 , the operations further comprising receiving, by the network device and from the network controller, an acceptable failure rate parameter for the time series algorithm, the acceptable failure rate parameter based at least in part on current local traffic conditions.

10. The system of claim 8 , the operations further comprising:

determining whether a current data point value collected is greater than a previous data point value;

based at least in part on determining that the current data point value is not greater than the previous data point value, continue providing the right-size capacity of the local resource for the network device; and

based at least in part of determining that the current data point value is greater than the previous data point value, execute a new time series calculation to determine a new right-size capacity of the local resource for the network device.

11. The system of claim 8 , wherein an individual time series algorithm of the set of times series algorithms includes training data based on historical consumption of the local resource for the network device.

12. The system of claim 8 , the operations further comprising uploading, by the network device and to the network controller, trained data for distribution to other network devices.

13. The system of claim 8 , the operations further comprising:

storing local time series conditions data in the network device that is collected prior to an event that approaches an unexpected exceeding of capacity of the local resource on the network device; and

providing the local time series conditions data as backpropagation training data for the set of time series algorithms for the network device.

14. The system of claim 8 , the operations further comprising including external indications of near future events for determining the right-size capacity of the local resource for the network device.

15. One or more non-transitory computer-readable media storing instructions that, when executed, cause one or more processors to perform operations comprising:

receiving, by a network device and from a network controller, a selection of a time series algorithm, for capacity right-sizing of a local resource, from a set of time series algorithms, wherein the time series algorithm is a computationally efficient algorithm based at least in part on current local traffic conditions and wherein a right-size capacity is at least an amount of the local resource needed to ensure proper network functioning;

receiving, by the network device and from the network controller, an instruction to run the time series algorithm when a local traffic condition trigger event occurs, wherein the instruction includes a determination of parameter values associated with the time series algorithm selection to be used in the time series algorithm;

receiving, by the network device and from the network controller, a determination of a number of data points for input to the time series algorithm, wherein data points are a sequence of values representing an amount of the local resource used by the network device at a point in time that are collected at predetermined time intervals;

determining, based at least in part on a calculation of the time series algorithm using the number of data points and parameter values and in response to the local traffic condition trigger event occurring, the right-size capacity of the local resource for the network device; and

providing, the right-size capacity of the local resource for the network device.

16. The one or more non-transitory computer-readable media of claim 15 , the operation further comprising receiving, by the network device and from the network controller, an acceptable failure rate parameter for the time series algorithm, the acceptable failure rate parameter based at least in part on current local traffic conditions.

17. The one or more non-transitory computer-readable media of claim 15 , the operation further comprising:

determining whether a current data point value collected is greater than a previous data point value;

based at least in part on determining that the current data point value is not greater than the previous data point value, continue providing the right-size capacity of the local resource for the network device; and

based at least in part of determining that the current data point value is greater than the previous data point value, execute a new time series calculation to determine a new right-size capacity of the local resource for the network device.

18. The one or more non-transitory computer-readable media of claim 15 , wherein an individual time series algorithm of the set of time series algorithms includes pre-provisioned training data.

19. The one or more non-transitory computer-readable media of claim 15 , the operation further comprising uploading, by the network device and to the network controller, trained data for distribution to other network devices.

20. The one or more non-transitory computer-readable media of claim 15 , the operation further comprising:

storing local time series conditions data in the network device that is collected prior to an unexpected exceeding of capacity of the local resource on the network device; and

providing the local time series conditions data as backpropagation training data for the set of time series algorithms for the network device.

Assignments (1)

Continuity (2)

References Cited (11)