US-20260128957-A1 - LARGE LEARNING MODEL FOR ACCESS POINTS

Abstract

Techniques to enable streamlining configuration and grouping of access points (APs) within a network. A Large Language Model (LLM) with Natural Language Processing (NLP) capabilities may be integrated into APs to allow administrators to use natural language for network management. The use of an LLM with NLP capabilities reduces the need for specialized technical knowledge to configure and to group APs, and, thus, simplifies AP grouping and configuration tasks. Such a context-aware system offers tailored recommendations for network optimization and simplifies the grouping and config of APs based on location.

Inventors

• Niloofar Bahadori
• Arman Rezaee
• Peiman Amini
• Juan Carlos Zuniga
• Jerome Henry

Assignees

• CISCO TECHNOLOGY, INC.

Dates

Publication Date

20260507

Application Date

20250812

Claims (20)

1. 1 . A method, comprising: receiving, at an access point (AP), one or more requests from a user; determining, using the AP, at least one processing action from the one or more requests using an Artificial Intelligence (AI) model configured for natural language interaction integrated into the AP; and causing, using the AP, the at least one processing action to be executed in real-time.
2. 2 . The method of claim 1 , wherein the at least one processing action includes updating a configuration of the AP and performing a troubleshooting of the AP.
3. 3 . The method of claim 1 , wherein data are stored on the AP and causing the at least one processing action to be executed in real-time includes executing the at least one processing action on the AP using the data.
4. 4 . The method of claim 1 , wherein causing the at least one processing action to be executed in real-time includes communicating, using the AP, with a neighboring AP.
5. 5 . The method of claim 4 , wherein communicating using the AP, with the neighboring AP includes communicating using an access switch that is in communication with the AP and the neighboring AP.
6. 6 . The method of claim 1 , wherein a type of the one or more requests is at least one selected from a group comprising a verbal request and a text request.
7. 7 . The method of claim 6 , wherein the Artificial Intelligence (AI) model configured for natural language interaction is at least one of a Large Language Model (LLM) and Small Language Model (SLM) that includes employing Natural Language Processing (NLP).
8. 8 . The method of claim 1 , wherein the one or more requests includes a verbal request, and wherein the user is in proximity to a location of the AP.
9. 9 . An access point (AP) system, comprising: a data storage configured to store data of the AP system; a communication component; and at least one processor that is integrated with an Artificial Intelligence (AI) model configured for natural language interaction, wherein the at least one processor is configured to: receive one or more requests from a user; determine at least one processing action from the one or more requests using the AI model configured for natural language interaction; and cause the at least one processing action to be executed in real-time.
10. 10 . The AP system of claim 9 , wherein the at least one processor is further configured to use data stored on the data storage of the AP system to execute the at least one processing action in real-time.
11. 11 . The AP system of claim 9 , wherein the at least one processing action includes updating a configuration of the AP system and performing a troubleshooting of the AP system.
12. 12 . The AP system of claim 9 , wherein the at least one processor is further configured to communicate with a neighboring AP system using an access switch.
13. 13 . The AP system of claim 9 , wherein a type of the one or more requests is at least one selected from a group comprising a verbal request and a text request.
14. 14 . The AP system of claim 13 , wherein the one or more requests includes the verbal request, and wherein the user is in proximity to a location of the AP system.
15. 15 . The AP system of claim 9 , wherein the AI model configured for natural language interaction is at least one of a Large Language Model (LLM) and a Small Language Model (SLM) that includes employing Natural Language Processing (NLP).
16. 16 . One or more non-transitory computer readable storage media encoded with instructions that, when executed by a computer processor of an access point (AP), cause the computer processor to perform operations including: receiving, at the AP, one or more requests from a user; determining, using the AP, at least one processing action from the one or more requests using an Artificial Intelligence (AI) model configured for natural language interaction integrated into the AP; and causing, using the AP, the at least one processing action to be executed in real-time.
17. 17 . The one or more non-transitory computer readable storage media of claim 16 , wherein the at least one processing action includes updating a configuration of the AP and performing a troubleshooting of the AP.
18. 18 . The one or more non-transitory computer readable storage media of claim 16 , wherein causing the at least one processing action to be executed in real-time includes communicating, using the AP, with a neighboring AP.
19. 19 . The one or more non-transitory computer readable storage media of claim 16 , wherein a type of the one or more requests is at least one selected from a group comprising a verbal request and a text request.
20. 20 . The one or more non-transitory computer readable storage media of claim 16 , wherein the AI model configured for natural language interaction is at least one of a Large Language Model (LLM) and a Small Language Model (SLM) including employment of Natural Language Processing (NLP).

Description

CROSS-REFERENCE TO RELATED APPLICATION This application claims priority to U.S. Provisional Patent Application No. 63/714,945, filed Nov. 1, 2024, the entirety of which is incorporated herein by reference. TECHNICAL FIELD The present disclosure relates to access points used in wireless networks. BACKGROUND In modern enterprise environments, the demand for robust, flexible, and secure wireless connectivity continues to increase. One key strategy to meet this demand involves configuring groups of wireless Access Points (APs) to work together seamlessly. This approach facilitates applications such as seamless roaming for constant connectivity across different areas, load balancing to manage the network efficiently under relatively heavy usage, and enhanced security policies uniformly applied across substantially all APs. For instance, hospitals require uninterrupted access to patient data as healthcare providers move throughout the facility, and conference venues need to distribute network load effectively during large events to maintain service quality. Additionally, retail chains and hotels generally offer differentiated access levels to staff and guests while managing network security and performance. The process of grouping and configuring network devices such as APs, access switches, routers, etc., presents several challenges. First, network administrators generally log into the controller for each AP and/or a cloud-based management platform to push configurations to the APs. Such a process of pushing configurations to APs may be time-consuming and prone to errors, especially in large deployments. Second, resources necessary to accurately identify the physical location of each AP within a network, and to effectively group APs with respect to a controller based on their locations or roles, are extensive. Meticulous planning and identifying physical locations of APs and grouping APs can be resource-intensive. Other challenges include scalability issues, configuration consistency, and performance optimizations. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagrammatic representation of a network that includes access points (APs), in accordance with an example embodiment. FIG. 2 is a block diagram representation of an AP, in accordance with an embodiment. FIG. 3 is a block diagram representation of a user device which may communicate with a network device such as an AP, in accordance with an embodiment. FIG. 4 is a flow chart depicting a method for causing a network device such as an AP to perform at least one processing action in response to one or more requests from a user, in accordance with an example embodiment. FIG. 5 is a diagrammatic representation of an AP in communication with another AP via an access switch, in accordance with an example embodiment. FIG. 6 is a flow chart depicting a method for causing an AP to communicate and perform at least one processing action on another AP in response to one or more requests from a user, in accordance with an example embodiment FIG. 7A is a diagrammatic representation of a user interface in which an access point provides information to a user, in accordance with an embodiment. FIG. 7B is a diagrammatic representation of a user interface in which an access point provides information to a user in response to a query in accordance with an embodiment. FIG. 8 illustrates a hardware block diagram of a computing device that may perform the functions of a mobile device, a client, a station, an access point, and/or a wireless local area network controller (WLC) referred to herein in connection with the techniques depicted in FIGS. 1-6, 7A and 7B. DESCRIPTION OF EXAMPLE EMBODIMENTS Overview Presented herein are techniques which enable a network device such as an access point (AP), access switch, or router to efficiently execute processing actions. A Large Language Model (LLM), as for example a LLM with Natural Language Processing (NLP) capabilities may be integrated substantially directly into an AP. When an LLM is integrated into an AP, the AP may be able to leverage the LLM without latency issues. According to one embodiment, methods are provided for causing an access point (AP) to execute at least one processing action. One or more requests from a user are received at an AP. The AP determines at least one processing action from the one or more requests using a Large language model integrated into the AP. In response to determining the at least one processing action, the AP causes the at least one processing action to be executed in real-time. In some aspects, the techniques described herein relate to a method, including: receiving, at an access point, one or more requests from a user; determining, using the AP, at least one processing action from the one or more requests using a Large Language Model integrated into the AP; and causing, using the AP, the at least one processing action to be executed in real-time. Example Embodiments Embodiments are presented herein