MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641118347 A) filed by A. Vasantharaj; A. Sharathi; P. Sundhar; and M. Parrin on October 03, 2026, for Intelligent 5g Network Slice Resource Optimization System Using Deep Reinforcement Learning.
Inventors include A. Vasantharaj; A. Sharathi; P. Sundhar; and M. Parrin.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: The present invention relates to an Intelligent 5G Network Slice Resource Optimization System Using Deep Reinforcement Learning for dynamically allocating shared network resources among multiple 5G network slices. The system comprises a network monitoring module configured to acquire traffic demand, resource utilization, latency, throughput, packet loss and resource availability information, and a slice requirement module configured to acquire slice-specific QoS, priority and SLA requirements. A state construction module combines the monitored information, slice requirements, resource availability and previous allocation information into a multidimensional network state. A deep reinforcement learning engine processes the network state and generates resource allocation actions for increasing, decreasing, maintaining or reallocating resources among network slices. A constraint verification module validates the generated actions against infrastructure capacity, slice-level resource limits and service requirements. A resource allocation module implements the validated actions across communication and computing resources. A performance evaluation module measures the resulting network performance, while a reward-generation mechanism produces feedback based on QoS, SLA compliance, resource utilization and allocation performance. The feedback is supplied to the deep reinforcement learning engine for subsequent optimization cycles. The invention thereby provides a continuous closed-loop mechanism for adaptive and intelligent resource optimization in heterogeneous 5G network-slicing environments. FIG.1.
Disclaimer: Curated by HT Syndication.
