๐ค AI Summary
To address the challenge of simultaneously achieving ultra-low latency and high reliability in 5G V2X Mode 2 communications under sparse and sporadic traffic conditions, this paper proposes a lightweight, distributed resource selection optimization method tailored for bursty safety-critical messages. Without relying on base station scheduling, the method redesigns the autonomous resource selection mechanism in Mode 2, with particular focus on interference mitigation and adaptive retransmission under random access. By modeling dynamic inter-vehicle channel competition, it incorporates local sensing enhancement and probabilistic backoff adjustment, ensuring low implementation complexity while significantly improving resource utilization efficiency. Simulation results demonstrate a 40% increase in system capacity, a 32% reduction in end-to-end latency, and a one-order-of-magnitude decrease in packet error rateโrobustly satisfying stringent road-safety requirements of <100 ms latency and 99.999% reliability.
๐ Abstract
The emerging road safety and autonomous vehicle applications require timely and reliable data delivery between vehicles and between vehicles and infrastructure. To satisfy this demand, 3GPP develops a 5G Vehicle-to-Everything (V2X) technology. Depending on the served traffic type, 5G V2X specifications propose two channel access methods: (i) Mode 1, according to which a base station allocates resources to users, and (ii) Mode 2, according to which users autonomously select resources for their transmissions. In the paper, we consider a scenario with sporadic traffic, e.g., a vehicle generates a packet at a random time moment when it detects a dangerous situation, which imposes strict requirements on delay and reliability. To satisfy strict delay requirements, vehicles use Mode 2. We analyze the performance of Mode 2 for sporadic traffic and propose several approaches to improve it. Simulation results show that the proposed approaches can increase the system capacity by up to 40% with a low impact on complexity.