China Communications • June 2021 ehicular networks are expected to empower auto mated driving and intelligent transportation via vehicle-to-everything (V2X) communications and edge/cloud-assisted computation, and in the meantime Cellular V2X (C-V2X) is gaining wide support from the global industrial ecosystem. The 5G NR-V2X technology is the evolution of LTE-V2X, which is expected to provide ultra-Reliable and Low-Latency Communications (uRLLC) with 1ms latency and 99.999% reliability.

more »

... ess, vehicular networks still face great challenges in supporting many emerging time-critical applications, which comprise sensing, communication and computation as closed-loops. On the one hand, compared with conventional mobile services, road and driving-related applications pose more strict latency requirements, beyond the air-interface delay, for road safety and efficiency. On the other hand, context information (e.g., the conditions of surrounding vehicles and pedestrians, dynamic high-precision maps, availability of parking lots, and traffic congestion) can be outdated due to high dynamics over various time scales, as measured by recently proposed new information timeliness/freshness metrics (e.g., age of information). In this regard, fundamental issues of time-critical communication (e.g., timely delivering the context information) and computation (e.g., making driving decisions with low response time) should be investigated to support emerging applications like intelligent connected automated driving. Inspired by these, this feature topic aims to bring new theories, frameworks, algorithms, mechanisms, applications, and tools of time-critical communications and computations for intelligent vehicular networks. The Call for Papers generated considerable erests in the research community, and after impartial and rigorous reviews, we have accepted eight papers covering the state-of-the-art work from transmission, scheduling, computing to applications. The feature topic begins with the article by Lu et al., "MARVEL: multi-agent reinforcement learning for VANET delay minimization". This article proposes a Multi-Agent Reinforcement Learning (MARL) based decentralized routing scheme, where the inherent similarity between the routing problem in vehicular ad hoc network and the MARL problem is exploited. The proposed routing scheme models the interaction between vehicles and the environment as a multi-agent problem in which each vehicle autonomously establishes the communication channel with a neighbor device regardless of the global information. Simulation performed in the 3GPP Manhattan mobility model demonstrates less than 45.8ms average latency and high stability of 0.05 % averaging failure rate with varying vehicle capacities. The article by Zhao et al., "Q-greedyUCB: a new exploration policy to learn resource-efficient scheduling," proposes a Reinforcement learning (RL) algorithm to find an optimal scheduling policy to minimize the delay for a given energy constraint in vehicular communication systems where the environments such as traffic arrival rates are not known in advance and can change over time. The problem is formulated as an infinite-horizon Constrained Markov Decision Process (CMDP). The Lagrangian relaxation technique and a variant of Q-learning, Q-greedyUCB that combines ε-greedy and Upper Confidence Bound (UCB) algorithms are used to solve the problem. The proposed Q-greedyUCB algorithm is proved to converge to an optimal solution and is thus superior to other existing methods. The algorithm can also learn and adapt to the environment changes, showing good robustness. The article by Zhu et al., "On latency reductions in