MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641080981 A) filed by Siddaganga Institute Of Technology on June 30, 2026, for System And Method For Decentralized Multi-Objective Task Offloading In Vehicular Edge Computing Networks.
Inventors include Dr. Keerthan Kumar T G; and Bhuvana K L.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: A system and method for decentralized, multi-objective computation task offloading in a vehicular edge computing (VEC) network. Each vehicle operates as an autonomous agent that, on a per-task basis, selects an execution mode among local on-board processing, vehicle-to-vehicle (V2V) offloading, and vehicle-to- infrastructure (V2I) offloading so as to maximize an individual utility jointly accounting for computation and communication delay, energy cost, offloading expense, task success, and service reliability, subject to per-task deadline, per-vehicle energy-budget, per-RSU capacity, NOMA power, transmit-power, and minimum-SINR constraints. The offloading interaction is modeled as a non-cooperative game solved by multi-agent deep reinforcement learning that drives the network toward a Nash equilibrium, and concurrent V2I uplinks are multiplexed by Non-Orthogonal Multiple Access (NOMA) with successive interference cancellation and congestion-based RSU pricing. The framework operates without a central controller, adapts to mobility, channel, and load variations, and substantially reduces latency while improving task success rate relative to baseline strategies.
Disclaimer: Curated by HT Syndication.
