MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202641079296 A) filed by Jayapandian Natarajan; Atmika M Banerjee; Grace Ann Mathew; and Dheenadayalan N on June 27, 2026, for Agentic Ai-Based Autonomous Dynamic Resource Management Framework With Adaptive Monitoring And Self-Healing Mechanism For Cloud Data Centers.
Inventors include Jayapandian Natarajan; Atmika M Banerjee; Grace Ann Mathew; and Dheenadayalan N.
The application for the patent was published on July 03, 2026, under issue no. 27/2026.
Abstract: The rapid growth of cloud computing has significantly increased the demand for efficient resource management mechanisms it is capable of supporting dynamic and large-scale workloads. The cloud data centers are required to manage continuously for changing workloads while maintaining service quality, resource efficiency, and operational sustainability. Existing scheduling techniques is mainly focus on reactive decision-making approaches. This will allocate the resources based on current system conditions. This will affect the result of resource utilization, increasing the response time, higher waiting time, Service Level Agreement (SLA) violations, and higher energy consumption. The traditional scheduling mechanisms are provide limited support for autonomous recovery model and predictive resource optimization. These limitations are to create a need for intelligent and self-managing cloud resource management. These solutions capable of adapting to dynamic workload conditions and improving overall cloud infrastructure efficiency. To address these limitations, the present invention proposes an Agentic AI-Based Autonomous Dynamic Resource Management Framework with Adaptive Monitoring and Self-Healing Mechanisms for Cloud Data Centers. The framework integrates an Adaptive Monitoring Agent (102), Prediction Agent (103), Allocation Agent (104), Self-Healing Agent (105), and Decision Knowledge Repository (106). The Adaptive Monitoring Agent dynamically regulates monitoring intervals based on workload behavior to improve monitoring efficiency and prediction accuracy. The Decision Knowledge Repository system maintains historical workload patterns, resource allocation results, migration records, recovery actions, and SLA violation information to support continuous learning and policy refinement. The monitoring agent collects real-time infrastructure metrics, while the prediction agent forecasts future workload demand using machine learning techniques including Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Transformer networks, or combinations thereof. Based on predicted workloads, the allocation agent proactively manages virtual machine provisioning, migration, consolidation, and deactivation. The self-healing agent is used to detects failures, abnormal resource utilization conditions, overloaded hosts, and potential SLA risks and performs corrective actions without human intervention. This experimental evaluation conducted using the CloudSim Plus simulation environment demonstrates improvements in response time, average waiting time, resource utilization, resource wastage, fault tolerance, and energy efficiency when compared with conventional scheduling methods (107). The invention provides an intelligent, adaptive, self-learning, reliable, and sustainable cloud resource management solution suitable for next-generation cloud data centers, hybrid cloud infrastructures, and large-scale distributed computing environments.
Disclaimer: Curated by HT Syndication.
