Deep Reinforcement Learning Enabled Energy-Efficient Resource Allocation in Energy Harvesting Aided V2X Communication

Song, Yuqian; Xiao, Yang; Chen, Yaozhi; Li, Guanyu; Li, Jun

doi:10.1109/pimrc54779.2022.9978099

Cited by 2 publications

(1 citation statement)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A two-phase fuzzy logic-based handoff scheme was proposed in [ 31 ], demonstrating its ability to reduce unnecessary handoffs and decision delays compared with conventional schemes. Also, the combination of deep reinforcement learning (RL) with resource allocation has been explored in [ 32 , 33 , 34 , 35 ]. Through the RL mechanism, parameter values and resource allocation are optimized based on the information learned from the environment, thereby enhancing communication performance.…”

Section: Related Workmentioning

confidence: 99%

Reuse Distance-Aided Resource Selection Mechanisms for NR-V2X Sidelink Communication

Yin,

Hwang

2023

Sensors

View full text Add to dashboard Cite

Cellular vehicle-to-everything (C-V2X) facilitates direct communication between vehicles and other user equipment (UE) to improve the efficiency of the Internet of vehicles communication through sidelink. In addition, in the new radio vehicle-to-everything (NR-V2X) Mode 2, users can automatically select resources using the conventional sensing-based semi-persistent scheduling (SB-SPS) resource selection algorithm. This mechanism allows users to generate a list of available resources after a sensing window, after which the users can randomly select resources, and the resource can be used continuously over multiple periods before reselection. However, during the sensing window, neighbors may generate a similar list of available resources, and random selection may lead to resource conflicts. This phenomenon may lead to deteriorated communication performance and increased latency due to incorrect reception. Therefore, this paper proposes a reuse distance-aided resource selection (RD-RS) method which integrates resource reuse distance judgement with SB-SPS to mitigate resource conflicts and interference caused by random selection. Moreover, the reuse distance judgement is performed before the final resource selection, and whether the user will select the current resource depends on the reuse distance between that user and other occupiers. Furthermore, the performance of the proposed scheme is compared with other algorithms. Simulation results show that the proposed RD-RS not only achieves a higher packet reception ratio (PRR) but also effectively reduces the inter-packet gap (IPG). Moreover, in specific scenarios, the proposed method outperforms conventional schemes by 9% in terms of PRR and 70% in terms of Range.

show abstract