Search

US-20260129100-A1 - SYSTEM AND METHOD FOR MANAGING AN ENHANCED PERFORMANCE NETWORK USING ROLE-BASED AGENTS

US20260129100A1US 20260129100 A1US20260129100 A1US 20260129100A1US-20260129100-A1

Abstract

Aspects of the subject disclosure may include, for example, a device, including: a processing system including a processor; and a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations of: training a role-based agent to handle queries from users of a network; training a role-based foundation model to retrieve data relevant to a role of a user; receiving a query from the user; and providing the query to the role-based agent, wherein the role-based agent uses an associated role-based foundation model to process the query, collect relevant data from the network, and formulate an answer to the query. Other embodiments are disclosed.

Inventors

  • Mark Stockert
  • Thomas J. Routt
  • Jerry Robinson
  • Vijay Bhaskar Uppala
  • Imad Benbrahim

Assignees

  • AT&T INTELLECTUAL PROPERTY I, L.P.
  • AT&T MOBILITY II LLC

Dates

Publication Date
20260507
Application Date
20260105

Claims (20)

  1. 1 . A device, comprising: a processing system including a processor; and a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations, the operations comprising: receiving a query from a role-based agent associated with a user of a network; validating a role of the user and access permissions to obtain a validated query; obtaining contextual information relevant to the query and the role of the user; routing the validated query and the contextual information to an orchestrator; receiving processed output from the orchestrator, the processed output generated by a role-based foundation model associated with the role of the user; and providing the processed output to the role-based agent to enable the role-based agent to formulate and provide a response to the user.
  2. 2 . The device of claim 1 , wherein the validating the role of the user and access permissions further comprises accessing a role-based domain manager that enforces role-based access controls (RBAC) to determine valid access for the query.
  3. 3 . The device of claim 1 , wherein the obtaining the contextual information relevant to the query and the role of the user includes collecting historical network performance data, traffic analytics, or user interaction behavior from a vector database.
  4. 4 . The device of claim 1 , wherein the processed output generated by the orchestrator includes summarized key performance indicators (KPIs) for a specific network slice relevant to the query.
  5. 5 . The device of claim 1 , wherein the providing of the processed output to the role-based agent includes formatting the response using a natural language generation mechanism of the role-based foundation model.
  6. 6 . The device of claim 1 , wherein the obtaining the contextual information further comprises accessing network slice data from a Network Slice Management Function (NSMF) to identify attributes of a network slice relevant to the query.
  7. 7 . The device of claim 1 , wherein the role-based foundation model applies supervised fine-tuning (SFT) using labeled datasets specific to network performance analysis and troubleshooting.
  8. 8 . The device of claim 1 , wherein the query from the role-based agent includes a request for additional bandwidth for a high-performance service application implemented in a 5G slice of a communications network.
  9. 9 . The device of claim 1 , wherein the providing of the processed output to the role-based agent further includes escalating the query to a different role-based agent with broader access permissions when the response cannot be formulated by the role-based agent.
  10. 10 . A non-transitory machine-readable medium, comprising executable instructions that, when executed by a processing system including a processor, facilitate performance of operations, the operations comprising: receiving, at an orchestrator, a query from a role-based agent associated with a user of a network; determining, at the orchestrator, one or more role-based foundation models relevant to the query based on a role of the user; invoking the one or more role-based foundation models to process the query and obtain processed output; incorporating contextual information, including network traffic data or historical network performance statistics, into the processed output; and transmitting the processed output and the contextual information to the role-based agent to formulate a response to the user.
  11. 11 . The non-transitory machine-readable medium of claim 10 , wherein the query includes a request to dynamically allocate additional resources to a network slice associated with the user.
  12. 12 . The non-transitory machine-readable medium of claim 10 , wherein the contextual information includes role-specific service usage patterns and user interaction behavior retrieved from a vector database.
  13. 13 . The non-transitory machine-readable medium of claim 10 , wherein the orchestrator uses a role-based domain manager to validate access permissions and select the one or more role-based foundation models.
  14. 14 . The non-transitory machine-readable medium of claim 10 , wherein the processed output incorporates predictive analytics performed by the one or more role-based foundation models to estimate future network performance or potential anomalies.
  15. 15 . The non-transitory machine-readable medium of claim 10 , wherein the operations further comprise dynamically instantiating virtual resources within a communications network to enhance performance based on the query.
  16. 16 . The non-transitory machine-readable medium of claim 10 , wherein the operations further comprise identifying a high-priority network slice associated with the user and integrating key performance indicators (KPIs) relevant to the high-priority network slice into the processed output.
  17. 17 . The non-transitory machine-readable medium of claim 10 , wherein the one or more role-based foundation models apply supervised fine-tuning based on labeled datasets for network troubleshooting and performance analysis.
  18. 18 . The non-transitory machine-readable medium of claim 10 , wherein the processed output includes recommendations for adjusting bandwidth or latency parameters to improve a high-performance application utilized by the user.
  19. 19 . A method of dynamically providing virtual resources by an orchestrator, comprising: receiving, by a processing system including a processor, a query from a role-based agent associated with a user of a network; validating, by the processing system, a role of the user and access permissions to obtain a validated query; obtaining, by the processing system, contextual information relevant to the validated query and the role of the user; routing, by the processing system, the validated query and the contextual information to an orchestrator; receiving, by the processing system, processed output from the orchestrator, the processed output generated by a role-based foundation model associated with the role of the user; and providing, by the processing system, the processed output to the role-based agent to enable the role-based agent to formulate and provide a response to the user.
  20. 20 . The method of claim 19 , comprising: providing, by the processing system, additional resources to a slice of the network responsive to the validated query.

Description

CROSS REFERENCE TO RELATED APPLICATIONS This application is a continuation of U.S. patent application Ser. No. 18/815,964 filed on Aug. 27, 2024. All sections of the aforementioned application are incorporated herein by reference in their entirety. FIELD OF THE DISCLOSURE The subject disclosure relates to a system and method for managing an enhanced performance network using role-based agents. BACKGROUND Cellular networks are very sophisticated in terms of volume of data traffic handled, data rates or speed, and low latency. Cellular networks are quite complex systems that enable mobile communication across vast geographic areas. They consist of a large number of cell sites, each with its own antenna and equipment, spread over a wide area. These sites are interconnected and managed by a central control system, which ensures seamless communication as users move from one location to another. Complexity arises from several factors, including frequency bands, network protocols and handover mechanisms. Cellular networks operate on specific frequency bands allocated by regulatory authorities. These bands are divided into channels, each assigned to a different cell to avoid interference. Rules and standards govern how data is transmitted and received over the network. Popular protocols include global system for mobile communications (GSM), code division multiple access (CDMA), and fourth generation (4G) long term evolution (LTE), each with its own advantages and limitations. When a mobile user moves from one cell to another, the network must transfer the connection seamlessly. This process, known as handover, ensures continuous service without dropped calls or interruptions. Despite their complexity, cellular networks are designed to be modular and expandable. The addition of new cells into the network is a fairly uncomplicated process, promoting modularity in expansion. This allows for the continuous growth and improvement of network coverage and capacity. Beyond 4G networks, network engineers and technical support teams manually perform significant portions of fifth generation (5G) and next generation (NG) network management and troubleshooting. This approach requires extensive knowledge and experience and is time-consuming and prone to human error. 5G networks introduced a new level of complexity, with a massive increase in the number of connected devices, higher data rates, and the implementation of advanced technologies such as network slicing and edge computing. This complexity makes manual management and troubleshooting not only more challenging but also time-consuming. Engineers must sift through vast amounts of data to diagnose issues, configure network parameters, and ensure optimal performance. The manual nature of these tasks increases the risk of human error, which can lead to network downtime, degraded service quality, or security vulnerabilities. The dynamic nature of 5G networks, characterized by their ability to adapt to changing traffic patterns and demands, requires constant vigilance and rapid response to maintain elevated levels of performance and reliability. Relying solely on manual processes can hinder the ability to respond swiftly to these changes, potentially impacting user experience and satisfaction. The time-consuming aspect of manual network management also has implications for operational efficiency and cost. Manual network management diverts skilled personnel from strategic tasks to routine maintenance, limiting the potential for innovation and improvement. As 5G networks continue to grow in size and complexity, the scalability of manual processes becomes increasingly unsustainable. In essence, while the expertise and judgment of human engineers are invaluable, the limitations of manual network management in the context of 5G highlight the need for more automated, intelligent systems. These systems can assist human operators, reduce the likelihood of errors, and ensure that the potential of 5G technology is fully realized in a reliable, efficient, and secure manner. Furthermore, the existing methods of network data analysis do not effectively leverage the potential of advanced AI techniques. AI has shown exceptional promise in various areas of technology, especially in data analysis and interpretation. However, applying AI in the context of next generation network management remains limited and unoptimized. Finally, the feedback loop between the human operators and the system is often disjointed in current systems. Continuous improvement of the system based on human feedback is crucial for the system's adaptability and effectiveness. Lack of a proper mechanism for collecting and incorporating human feedback into the system limits potential improvements and adaptability of the system. BRIEF DESCRIPTION OF THE DRAWINGS Reference will now be made to the accompanying drawings, which are not necessarily drawn to scale, and wherein: FIG. 1 is a block diagram illustrating an exemplary, no