MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641110300 A) filed by Vellore Institute Of Technology on September 14, 2026, for A Unified Predictive And Energy-Aware Cloud Resource Orchestration System With Sla-Adaptive Multi-Metric Optimization.
Inventors include Anny Leema; P Balakrishnan; Asjad Iftakhar; Ayush Kumar; Ashish Sharma; and Tapan Jaydeo Wankhade.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: Thepresent invention relates to aunified predictiveand energy-awarecloud resourceorchestration system (100) for distributed cloud infrastructures. The system (100) integrates a hybrid virtualization selection engine for selecting between Virtual Machine, Container, and Serverless environments via cost-latency trade-off analysis; a multi-metric load modeling module computing weighted composite load scores from CPU, memory, and bandwidth utilization; and a predictive load estimation module applying exponential smoothing-based forecasting for proactive overload detection. An SLA-adaptive dynamic threshold engine dynamically recalibrates migration thresholds based on real-time SLA violation rates, replacing static threshold policies. An energy-aware consolidation module employs a linear power consumption model differentiating active and idle states, quantifying measurable energy savings against a full-load baseline. Experimental validation using CloudSim 3.0 demonstrates 43.88% cost reduction, 49.64% latency reduction, and 40%energyreduction compared to baselineVM-only static deployment, confirming ascalable and technically robust next-generation cloud resource management solution.
Disclaimer: Curated by HT Syndication.
