MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641091033 A) filed by Srm Institute Of Science And Technology, Ramapuram Campus on July 27, 2026, for System And Method For Semantic Relationship Propagation Analysis In Multi-Generation Scientific Citation Knowledge Graphs.
Inventors include V. Akash Sujith; Dr. R. Krishna Kumari; and Dr. K. Janaki.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: Abstract: Optimizing sensor activation to maximize coverage, network longevity, and network connectivity remains a fundamental challenge in deploying Wireless Sensor Networks (WSNs) across complex, unstructured environments. This paper introduces a novel, decentralized reinforcement learning framework based on Asynchronous Cellular Learning Automata (ACLA) for the autonomous deployment of sensors over three-dimensional terrains by incorporating Digital Elevation Model (DEM). We discretize the terrain as a lattice, where each node embodies a sensor equipped with a Learning Automaton with adaptive learning schemes. These agents independently and intelligently adjust their active/sleep schedules based purely on adaptive rules for localized feedback as reward or penalty, derived from their immediate topographic features, such as elevation and gradient. The core of our investigation is to evaluate their efficacy in achieving a critical multi-objective optimization: maximizing the Coverage Volume Ratio (CVR), Network connectivity Ratio (CR), while also maximizing the Energy Efficiency Index (EEI). This approach helps the sensor nodes to update their schedules asynchronously through independent Poisson clocks, which do not require global synchronization. Our simulation results on Digital Elevation Model-based terrains demonstrate that the proposed ACLA-based framework enables a self-organizing WSN that achieves optimal coverage with minimal energy consumption and full network connectivity compared to fixed learning rate approaches. to dynamically and efficiently approximate an optimal coverage configuration. The findings substantiate that asynchronous ACLA mechanisms with adaptive learning and DEM-guided connectivity recovery provide a robust, scalable, and infrastructure-free strategy for terrain-aware sensor management, paving the way for more resilient and energy-sustainable deployments in real-world uneven environments.
Disclaimer: Curated by HT Syndication.
