Intellectual Property India Publishes Patent Application for 'Dual Stage Deep Reinforcement Learning For Cloud Task Scheduling' Filed by Vellore Institute Of Technology

MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541123285 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on Dec. 6, 2025, for 'dual stage deep reinforcement learning for cloud task scheduling.'

Inventor(s) include Ujjwal Shivchand Daharwal; and Sumit Kumar Jindal.

The application for the patent was published on Jan. 2, under issue no. 01/2026.

According to the abstract released by the Intellectual Property India: "The present disclosure provides a computer-implemented method (100) for dual-stage deep reinforcement learning task scheduling in edge computing environments. The method includes initializing (102) an edge computing environment (300) with server farms (302, 316) containing edge servers (304, 310, 318) hosting virtual machines (306, 308, 312, 314, 320, 322). The method generates (104) a directed acyclic graph of computational tasks with dependency relationships and creates (106) a task ready queue (334) containing tasks eligible for scheduling. The method executes (110) a first stage deep reinforcement learning agent (326) using policy-based reinforcement learning for task prioritization based on deadlines and dependency constraints, and executes (114) a second stage agent (328) using value-based reinforcement learning for resource allocation. The method coordinates both agents and updates (118, 122) system state based on execution results."

Disclaimer: Curated by HT Syndication.