MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641117799 A) filed by Valli Priyadharshini K; Burra Anupama; Dr. Shahin Fatima; Chandrakala V; G. Mahalakshmi; Dr. B. Chitradevi; Keerthana M; R. Deepa; Dr. N. Kalyani; and M. Ranjani on October 01, 2026, for Carbon-Aware Adaptive Task Orchestration System For Energy-Efficient Edge Iot Computing.
Inventors include Valli Priyadharshini K; Burra Anupama; Dr. Shahin Fatima; Chandrakala V; G. Mahalakshmi; Dr. B. Chitradevi; Keerthana M; R. Deepa; Dr. N. Kalyani; and M. Ranjani.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: ABSTRACT The proposed Carbon-Aware Adaptive Task Orchestration System for Energy-Efficient Edge IoT Computing provides an intelligent framework for reducing energy consumption and carbon emissions during IoT task execution across heterogeneous edge devices. The system continuously monitors CPU utilization, memory usage, network traffic, battery level, execution latency, power consumption, and carbon-intensity information associated with the edge environment. A Long Short-Term Memory (LSTM) model is employed to predict short-term workload and carbon-intensity variations. Based on these predictions, a Multi-Objective Particle Swarm Optimization (MOPSO) algorithm dynamically determines task placement and scheduling across edge nodes and cloud resources. The optimization simultaneously minimizes energy consumption, carbon emissions, execution latency, and workload imbalance while satisfying task deadlines. A Dynamic Voltage and Frequency Scaling (DVFS) technique is incorporated to adjust processor frequency according to workload requirements, while Deep Q-Network (DQN)-based adaptive offloading determines whether computational tasks should be executed locally, transferred to a neighboring edge node, or offloaded to the cloud. A carbonaware task- priority mechanism further assigns higher priority to execution opportunities associated with lower predicted carbon intensity. The proposed system therefore combines LSTM-based prediction, MOPSO-based multi-objective task orchestration, DQN-based computation offloading, and DVFS-based energy optimization to achieve adaptive and carbonefficient IoT computing. The framework is designed for resource-constrained edge environments where computational efficiency, latency, energy consumption, and environmental sustainability must be jointly optimized.
Disclaimer: Curated by HT Syndication.
