Multi-agent transformer approach for collaborative task offloading and resource optimization in NOMA-based vehicular edge computing

doi:10.1016/j.adhoc.2026.104222

Multi-agent transformer approach for collaborative task offloading and resource optimization in NOMA-based vehicular edge computing

2026-03-14

0

PRE

AI

DOI：10.1016/j.adhoc.2026.104222

Original

Original request for help

Save

Abstract

En 中文

• Proposed NOMA-based task offloading and resource optimization scheme for vehicular edge computing with enhanced processing efficiency and system reliability. • Dual optimization framework decomposing the problem into MAT reinforcement learning for offloading decisions and Lagrangian optimization for resource allocation. • Pioneering application of multi-agent Transformer algorithm to vehicular edge computing environments with sequential decision mechanisms and collaborative learning. • Simulation validation using real vehicle trajectory data demonstrating superior performance in offloading success rates, resource utilization, and processing times.

Keywords:

NOMA

vehicular edge computing

task offloading

resource optimization

multi-agent Transformer

Multi-agent transformer approach for collaborative task offloading and resource optimization in NOMA-based vehicular edge computing

Abstract

Journal

Organization