马忠贵, 李卓, 梁彦鹏. 自动驾驶车联网中通感算融合研究综述与展望[J]. 工程科学学报, 2023, 45(1): 137-149. DOI: 10.13374/j.issn2095-9389.2022.04.16.003
引用本文: 马忠贵, 李卓, 梁彦鹏. 自动驾驶车联网中通感算融合研究综述与展望[J]. 工程科学学报, 2023, 45(1): 137-149. DOI: 10.13374/j.issn2095-9389.2022.04.16.003
MA Zhong-gui, LI Zhuo, LIANG Yan-peng. Overview and prospect of communication-sensing-computing integration for autonomous driving in the internet of vehicles[J]. Chinese Journal of Engineering, 2023, 45(1): 137-149. DOI: 10.13374/j.issn2095-9389.2022.04.16.003
Citation: MA Zhong-gui, LI Zhuo, LIANG Yan-peng. Overview and prospect of communication-sensing-computing integration for autonomous driving in the internet of vehicles[J]. Chinese Journal of Engineering, 2023, 45(1): 137-149. DOI: 10.13374/j.issn2095-9389.2022.04.16.003

自动驾驶车联网中通感算融合研究综述与展望

Overview and prospect of communication-sensing-computing integration for autonomous driving in the internet of vehicles

  • 摘要: 为了应对自动驾驶车联网极低的通信时延、极高的可靠性、更高的传输速率等极致性能需求,亟需破解现有车联网中通信、感知、计算相互割裂与独立分治的问题,实现“云−边−端”一体化协同感知、协同传输和协同决策。为此,急需对自动驾驶车联网的通感算融合开展研究,实现三者的高效融合。首先论述了目前在通信、感知、计算融合领域的研究进展,然后给出了通感算融合网络的定义,论述了通感助算、通算助感以及感算助通的研究进展。针对自动驾驶车联网的应用场景,创造性地提出了“五层四面”通感算融合的网络架构,横向五层自下而上分别是:多元接入层、统一网络层、多域资源层、协同服务层、管理与应用层;纵向四面分别是:通信面、感知面、算力面、智能融合面,通过五层四面的深度融合,进一步提升了自动驾驶车联网中通感算融合网络的性能。其次,提出了评价通感算融合网络的性能指标体系,最后针对目前研究存在的问题以及未来发展方向给出了四点可行性建议。

     

    Abstract: To meet extreme performance requirements, such as extremely low communication delay, extremely high reliability, and a higher transmission rate, for autonomous driving in the Internet of vehicles (IoV), the future IoV should be merged into a united framework that integrates communication, sensing, and computing. At the same time, as the 5G-Advanced system moves toward supporting a broader toB vertical industry, it will face a more complex and heterogeneous user environment and multidimensional digital space, which requires 5G-Advanced terminals and 5G-Advanced networks to have stronger environmental sensing, computing, and intelligence capabilities. To realize deep integration for autonomous driving in the IoV, the sensing of IoV depends on not only radar positioning, camera imaging, and various sensor detections but also communication, which can collect a variety of data to the edge node for calculation. At the same time, with the support of cloud edge and end integration efficient computing power to achieve high-precision sensing and efficient communication, the integration network further improves collaborative mobile computing robustness. Therefore, the three functions of communication, sensing, and computing for autonomous driving in the IoV are interrelated and promote each other. To break through the architectural barrier of universal sensing integration in the Internet of autonomous vehicles, it is necessary to explore how to build a universal sensing integration network architecture with decoupled resources, scalable capabilities, and reconfigurable architecture, as well as universal sensing integration resource management technology. However, communication, sensing, and computing are separated from each other in the existing IoV. Thus, we scrutinize the scientific problem of the endogenous integration of communication, sensing, and computing for autonomous driving in the IoV. First, the current research progress in integrating communication, sensing, and computing is discussed. Second, communication-sensing-computing-integrated IoV is defined, and the research progress on communication-sensing-assisted computing, communication-computing-assisted sensing, and sensing-computing-assisted communication is discussed. Aiming at the scenario of an IoV for autonomous driving, the architecture of communication-sensing-computing-integrated IoV with five layers and four planes is proposed. The horizontal five layers from bottom to top are a multiple access layer, unified network layer, multi-domain resource layer, collaborative service layer, and management and application layer. The four vertical planes are communication, sensing, computing power, and intelligent integration planes, respectively. Deeply integrating the five layers and four planes further improves the performance of the integrated IoV. Third, key performance indexes for evaluating the integrated IoV are proposed. Finally, four feasible suggestions are given for the current research problems and the future development direction.

     

/

返回文章
返回