Treffer: 基于IPv6+的智能车联算网调度方案设计与实现.

To address the challenges of high infrastructure costs, poor IPv4 network programmability, and insufficient resource scheduling flexibility in Internet of vehicles, an IPv6+ based intelligent computing-network scheduling scheme was proposed. A modular design encompassing computing-network access, awareness, and scheduling was employed to achieve precise perception and efficient coordination of "end-edge-cloud" computing, network and application. The scheduling module constructed scenario-specific strategies based on atomic orchestration, leveraging the A2C reinforcement learning algorithm to dynamically optimize task offloading locations and computing frequencies, thereby balancing energy consumption and latency. Furthermore, it integrated SRv6 programmable routing technology to autonomously select optimal transmission paths, realizing IPv6+ based computing-network synergistic optimization. Simulation results demonstrate the effectiveness and superiority of the A2C algorithm in supporting multi-objective optimization. Field deployments validate the feasibility and practicality of the proposed scheme in enabling flexible resource scheduling and reducing computing infrastructure costs, while providing a solution with both theoretical innovation and engineering value for constructing collaborative vehicle-road-cloud systems. [ABSTRACT FROM AUTHOR]

针对车联网中算力基础设施成本高、IPv4 网络可编程性差、资源调度灵活性不足的问题, 提出了基于 IPv6+的智能车联算网调度方案。通过算网接入、算网感知和算网调度的模块化设计, 实现对"端-边-云"算 力、网络及应用的精准感知与高效调度。算网调度模块基于原子编排构建场景化调度策略, 利用强化学习A2C 算法动态优化任务卸载位置与算力主频, 权衡能耗与时延。同时, 结合SRv6 可编程路由技术自主选择最优传输 路径, 实现基于IPv6+的算网协同优化。仿真实验表明了A2C算法支持系统多目标优化的有效性和优越性, 实地 部署验证了所提方案在资源灵活调度和降低算力基础设施成本方面的可行性与实用性, 为车路云协同体系建设 提供了兼具理论创新与工程价值的解决方案. [ABSTRACT FROM AUTHOR]

Copyright of Journal on Communication / Tongxin Xuebao is the property of Journal on Communications Editorial Office and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)