AUTOMATED NETWORK SLICING IN 6G NETWORKS USING REINFORCEMENT LEARNING

Abstract

This invention provides a method for automated network slicing in 6G networks using reinforcement learning (RL) to dynamically manage resource allocation across diverse network slices. The system includes an RL agent that learns optimal policies for distributing resources to meet Quality of Service (QoS) requirements, including latency, throughput, reliability, and energy efficiency. The RL agent operates through real-time feedback from a QoS monitoring system, adjusting network configurations based on observed performance and demand patterns. In one embodiment, a single-agent framework manages network slices centrally; in another, a hierarchical multi-agent framework handles complex, heterogeneous environments with a central coordinating agent. An edge-distributed RL framework enables low-latency, localized decision-making. The invention offers a scalable, adaptive solution for 6G network resource management, ensuring consistent QoS while reducing operational costs and improving overall network efficiency.

Specification

Description:The embodiments of the present invention generally relates to the field of 6G wireless communication networks, focusing specifically on the automated management of network resources through network slicing. More particularly, it involves the application of reinforcement learning techniques to dynamically and intelligently allocate resources in a 6G network environment. This ensures optimal Quality of Service (QoS) for diverse applications and services, adapting in real time to fluctuating network conditions.
BACKGROUND OF THE INVENTION
The following description of related art is intended to provide background information pertaining to the field of the disclosure. This section may include certain aspects of the art that may be related to various features of the present disclosure. However, it should be appreciated that this section be used only to enhance the understanding of the reader with respect to the present disclosure, and not as admissions of prior art.

With the evolution of wireless technology, each , Claims:1. A method for automated network slicing in a 6G network, comprising:
initializing a plurality of network slices within a 6G network environment, each slice configured with initial resource parameters based on anticipated application requirements;
deploying a reinforcement learning (RL) agent that interacts with the network environment, wherein the RL agent receives feedback on Quality of Service (QoS) parameters from each network slice;
dynamically adjusting resource allocation across the plurality of network slices based on policies learned by the RL agent to optimize one or more QoS parameters;
continuously updating the policies of the RL agent based on real-time feedback from the network environment, wherein the RL agent rewards resource configurations that improve QoS metrics and penalizes those that degrade QoS performance.

2. The method of claim 1, wherein the RL agent dynamically allocates resources based on at least one QoS metric selected from latency, throughput, reliability, and energy efficienc

Documents