Optimizing key allocation during roaming using machine learning

Patent metadata

Field	Value
Publication number	US-12389285-B2
Application number	US-202418410839-A
Country	US
Kind code	B2
Filing date	Jan 11, 2024
Priority date	Aug 2, 2021
Publication date	Aug 12, 2025
Grant date	Aug 12, 2025

Abstract

Systems and methods are provided for optimizing resource consumption by bringing intelligence to the key allocation process for fast roaming. Specifically, embodiments of the disclosed technology use machine learning to predict which AP a wireless client device will migrate to next. In some embodiments, machine learning may also be used to select a subset of top neighbors from a neighborhood list. Thus, instead of allocating keys for each of the APs on the neighborhood list, key allocation may be limited to the predicted next AP, and the subset of top neighbors. In some embodiments, a reinforcement learning model may be used to dynamically adjust the size of the subset in order to optimize resources while satisfying variable client demand.

First claim

What is claimed is:

1. A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to perform a method comprising:
learning migration behavior associated with a client device including patterns of migration amongst access points (APs) and to which APs the client device connects most often;
receiving a notification that comprises a current and a neighborhood list, wherein the current AP is the AP to which the client device is currently connected, and the neighborhood list is a list of APs in the neighborhood of the current AP;
predicting an AP in the neighborhood of the current AP to which the client device will migrate next based on predicted probabilities of connections to the APs in the neighborhood, the predicted probabilities being based on the current AP and the learned migration behavior;
generating an authentication client key for the predicted next AP; and
propagating the authentication client key to the predicted next AP to facilitate pre-authentication of the client device prior to migration of the client device to the predicted next AP.

2. The non-transitory computer-readable storage medium of claim 1 , wherein the authentication client key is a cryptographic key associated with the client device.

3. The non-transitory storage medium of claim 1 , wherein the client device one of skips or undergoes shortened authentication procedures upon attempting to associate to the predicted next AP, the authentication client key having been cached by the predicted next AP.

4. The non-transitory storage medium of claim 1 , wherein the instructions that when executed cause the learning of the migration behavior comprise further instructions causing the computing system to track movement of the client device between APs over time.

5. The non-transitory storage medium of claim 4 , wherein the instructions that when executed cause the learning of the migration behavior comprise further instructions causing the computing system to record to which AP the client device is connected at various points over the time during which the movement of the client device is tracked.

6. The non-transitory computer-readable storage medium of claim 1 including further instructions that when executed by the at least one processor of the computing system, causes the computing system to derive the neighborhood list based on path loss value relative to the current AP.

7. The non-transitory computer-readable storage medium of claim 6 , wherein the instructions that when executed cause the derivation of the neighborhood list comprise further instructions to cause the computing system to calculate path loss values between the current AP and APs neighboring the current AP.

8. The non-transitory computer-readable storage medium of claim 7 , wherein the instructions that when executed cause the predicting of the AP to which the client device will migrate next comprise further instructions to cause the computing system to execute one or more machine learning operations to determine a probability of being an AP to which the client device will migrate next.

9. The non-transitory computer-readable storage medium of claim 8 , wherein the instructions that when executed cause the predicting of the AP to which the client device will migrate next comprise further instructions to cause the computing system to select that AP having the highest probability of being the AP to which the client device will migrate next.

10. A system, comprising:
at least one processor; and
a memory storing instructions that, when executed by the at least one processor, cause the system to perform a method comprising:
learning migration behavior of a client device including patterns of migration amongst access points (APs) and to which APs the client device connects most often;
determining a current AP to which the client device is currently connected
allocating a cryptographic authentication key associated with the client device to a predicted target AP to which the client is expected to migrate, the predicted target AP having been determined based on predicted probabilities of connections to the APs in a neighborhood list, the predicted probabilities being based on the current AP and the learned migration behavior of the client device; and
reestablishing at least one of existing security or quality of service (Qos) parameters relevant to the client device prior to reconnecting to the predicted target AP.

11. The system of claim 10 , wherein the instructions that when executed cause the system to learn the migration behavior comprise further instructions that cause the system to track movement of the client device between APs over time.

12. The system of claim 11 , wherein the instructions that when executed cause the learning of the migration behavior comprise further instructions that cause the system to record to which AP the client device is connected at various points over the time during which the movement of the client device is tracked.

13. The system of claim 10 , wherein the memory stores further instructions that when executed by the at least one processor of the computing system, causes the system to derive the neighborhood list based on path loss value relative to the current AP.

14. The system of claim 13 , wherein the instructions that when executed cause the derivation of the neighborhood list comprise further instructions that cause the system to calculate path loss values between the current AP and APs neighboring the current AP.

15. The system of claim 10 , wherein the instructions that when executed cause the predicting of the target AP to which the client device will migrate next comprise further instructions to cause the computing system to execute one or more machine learning operations to determine a probability of being an AP to which the client device will migrate next.

16. The system of claim 15 , wherein the instructions that when executed cause the predicting of the target AP to which the client device will migrate next comprise further instructions to cause the computing system to select that AP having the highest probability of being the AP to which the client device will migrate next.

17. The method of claim 13 , wherein the instructions that when executed cause the system to learn the migration behavior comprise further instructions that cause the system to build a Markov chain where each connection point of the Markov chain represents an AP on neighborhood list.

18. The system of claim 10 , wherein the client device one of skips or undergoes shortened authentication procedures upon attempting to associate to the predicted target AP, prior to connection of the client device to the predicted target AP, the cryptographic authentication key having been cached by the predicted target AP.

19. A method, comprising:
receiving at a target access point (AP) predicted to be an AP to which a client device will connect, a cryptographic authentication key indicative of the client device, the predicted target AP having been identified based on learned patterns of migration amongst APs and to which APs the client device connects most often;
caching at the target AP, the cryptographic key, prior to a BSS transition;
upon the client device associating to the target AP, pre-authenticating the client device by reestablishing at least one of existing security or quality of service (QOS) parameters relevant to the client device at the target AP; and
after associating the client d

Assignees

• Hewlett Packard Entpr Dev Lp

Inventors

• Gupta Gopal
• Mishra Abhinesh
• Theogaraj Isaac
• Ganu Sachin
• Bandemer Bernd
• Tellado Jose

Classifications

• H04W36/00835

  Determination of neighbour cell lists · CPC title

• H04W36/0061

  of neighbour cell information · CPC title

• H04W88/08

  Access point devices · CPC title

• H04W12/041

  Key generation or derivation · CPC title

• G06N20/00

  Machine learning · CPC title