MUMBAI, India, Feb. 6 -- Intellectual Property India has published a patent application (202541126585 A) filed by Vardhaman College Of Engineering, Ranga Reddy, Telangana, on Dec. 14, 2025, for 'intelligent spectrum management using deep reinforcement learning.'
Inventor(s) include Dr. Muniganti Radha Krishna; Mr. Madasu Ashwan Kumar; Ms. Kolarkar Ashlesha Vijay; Ms. Monika Garg; Ms. G Anusha; and Ms. R Arul Selvi.
The application for the patent was published on Feb. 6, under issue no. 06/2026.
According to the abstract released by the Intellectual Property India: "Intelligent Spectrum Management Using Deep Reinforcement Learning is the proposed invention. Deep reinforcement learning is proposed to address the challenges of inefficient spectrum utilization in modern wireless communication networks. The proposed approach enables autonomous and dynamic allocation of radio frequency resources by learning optimal spectrum access and control policies through continuous interaction with the wireless environment. Real-time spectrum sensing data, including channel availability, interference levels, and traffic demand, are used as input to a deep reinforcement learning agent that determines actions such as channel selection, bandwidth allocation, and power control. A reward-driven learning mechanism optimizes key performance metrics including spectral efficiency, throughput, latency, and fairness. The system operates without requiring explicit mathematical models of the radio environment, allowing effective adaptation to highly dynamic and uncertain conditions. The proposed intelligent spectrum management framework is well suited for cognitive radio networks, dense IoT deployments, and next-generation wireless systems, offering improved efficiency, scalability, and reliability."
Disclaimer: Curated by HT Syndication.
