Decentralized Resource Allocation-Based Multiagent Deep Learning in Vehicular Network

Resource allocation (RA) has a significant impact on vehicular network performance. With high mobility, RA is more challenging, as the number of vehicles in close proximity changes dynamically nonstationary environment. In this article, we propose multiagent double deep Q-networks scheme to stabilize system and maximize sum-capacity vehicle-to-infrastructure (V2I) links, while satisfying reliability delay constraints for vehicle-to-vehicle (V2V) links. To avoid interference caused by unstable V2V transmission mode selection considered design. addition, introduce binarized weight algorithm accelerate neural learning process and, therefore, improve computational complexity our scheme. Through extensive simulations analysis, demonstrate that proposed yields excellent performance terms sum-rate probability rate V2I communication modes. We also compare with weights other algorithms accuracy evaluation.