MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202641074359 A) filed by Vellore Institute Of Technology on June 16, 2026, for Reinforcement Learning-Based Trust-Aware Routing For Mobile Ad Hoc Networks.
Inventors include Anindita Kundu; and Lekha S.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: ABSTRACT TRUST-AWARE ROUTING FOR MOBILE AD HOC NETWORKS USING REINFORCEMENT LEARNING A computer-implemented method for routing in a Mobile Ad Hoc Network (MANET) comprises generating, by a processor (100), a dynamic network of mobile nodes (102) characterized by position, mobility, and residual energy, and establishing communication links (104) based on a transmission range. The method comprises monitoring packet forwarding behavior, computing a packet forwarding ratio (106), and computing a trust value (108) using a weighted moving average combining a historical trust value and the packet forwarding ratio (106). The method comprises identifying unreliable mobile nodes (102), computing a mobility stability factor (110), computing a routing success probability (112), maintaining a trust history buffer (114), and computing a predicted trust value (116) using a weighted temporal model. A network state (118) is modeled as a tuple, and a reinforcement learning agent (120) selects routing actions and updates a Q-table (122) to determine an optimal packet forwarding path.
Disclaimer: Curated by HT Syndication.
