安全增强的车载嵌入式系统中任务映射和调度算法

Task Mapping and Scheduling Algorithms for Security-enhanced Automotive Embedded Sys- tems

摘要

随着现代车载嵌入式系统应用的不断集成与电子控制单元（ECU）数量的增长，系统实时性保障面临更大挑战。同时，汽车与外部环境交互频繁，具有灵活数据速率的控制器局域网（CAN FD）虽提升了传输性能，但仍缺乏内置安全机制，易受伪装等攻击威胁。而添加安全机制往往会占用实时性资源，危害车辆安全，因此必须在保证系统实时性的同时提升安全性。为此，本文提出了一-种基于强化学习的任务咉射和调度算法（Rcinforccmcnt Lcarning-bascd Task Mapping and Scheduling Algorithm， RLMS），将任务映射建模为马尔科夫决策过程，结合资源感知机制以 ECU 利用率为约束优化方案，在满足实时性约束的前提下减少 CAN FD 总线消息数量，并为消息提供 4 字节 MAC 的基础安全保护。为进一步提升系统安全性，设计安全增强机制 （Security Enhancement with Balanced Rounds，SEBR），利用系统空闲时间逐轮扩展消息的MAC 字节。最终，通过真实案例和模拟实验验证了所提方法的有效性。

Abstract

As modern in-vehicle embedded systems continue to integrate more applications and the number of Electronic Control Units （ECUs）increases，ensuring system real-time performance becomes increasingly challenging.Meanwhile，frequent interactions between vehicles and the external environment raise security concerns.Although the Controller Area Network with Flexible Data-Rate（ CAN FD）improves transmission performance through flexible data rates，it lacks built-in security mechanisms and remains vulnerable to threats such as spoofing.Introducing security mechanisms often consumes real-time resources and may compromise vehicle safety. Therefore，enhancing security while ensuring real-time guarantees is essential.This paper proposes a Reinforcement Learning-based Task Mapping and Scheduling Algorithm（RLMS），which formulates the task mapping process as a Markov Decision Process.By in- corporating a resource-aware mechanism that constrains ECU utilization，the proposed method reduces the number of CAN FD bus messages while satisfying real-time constraints.Each message is provided with basic security protection through a 4 -byte Message Au- thentication Code（MAC）.To further strengthen system security，a mechanism called Security Enhancement with Balanced Rounds （SEBR）is introduced，which gradually increases the MAC length by leveraging system idle time in a round-by-round manner.The ef- fectiveness of the proposed approach is validated through real-world case studies and simulation experiments.

关键词

嵌入式系统 / CAN FD / 任务调度 / 安全性增强

Key words

embedded systems / CAN FD / task scheduling / security enhancement

参考文献

[1] Zou A, Li J, Gill C D, et al. RTGPU:real-time GPU scheduling of hard deadline parallel tasks with fine-grain utilization[J]. IEEE Transactions on Parallel and Distributed Systems, 2023, 34 (5): 14501465.

[2] Sonko S, Daudu C D, Osasona F, et al. The evolution of embedded systems in automotive industry:a global review[J]. World Journal of Advanced Research and Reviews, 2024, 21(2):96-104.

[3] Sun H, Huang W, Weng J, et al. CCID-CAN:cross-chain intrusion detection on CAN bus for autonomous vehicles[J]. IEEE Internet of Things Journal, 2024, 11(15):26146-26159.

[4] de Andrade R, Santos M M D, Justo J F, et al. Security architecture for automotive communication networks with CAN FD[J]. Com- puters ＆ Security, 2023,129:103203,doi:10.1016/j.cose.2023.103203.

[5] Lodge N, Tambe N, Saqib F. Addressing vulnerabilities in CAN- FD:an exploration and security enhancement approach[J]. Internet of Things, 2024, 5(2):290-310.

[6] Shang C, Cao J, Liu J, et al. CEAMP:a cross-domain entity authen- tication and message protection framework for intra-vehicle network[J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 25(7):6780-6795.

[7] Zhao R, Luo C, Gao F, et al. Application-layer anomaly detection leveraging time series physical semantics in can fd vehicle networks[J]. Electronics, 2024, 13 ( 2 ):377,doi: 10.3390/electron-ics13020377.

[8] Xie Y, Zeng G, Kurachi R, et al. Security/timing-aware design space exploration of CAN FD for automotive cyber-physical sys- tems[J]. IEEE Transactions on Industrial Informatics, 2019, 15 (2):1094-1104.

[9] Roy S K, Devaraj R, Sarkar A. Contention cognizant scheduling of task graphs on shared bus-based heterogeneous platforms[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2021, 41(2):281-293.

[10] Minaeva A, Roy D, Akesson B, et al. Control performance optimi- zation for application integration on automotive architectures[.I] IEEE Transactions on Computers, 2020, 70(7):1059-1073.

[11] McLean S D, Juul Hansen E A, Pop P, et al. Configuring ADAS platforms for automotive applications using metaheuristics[J]. Frontiers in Robotics and AI, 2022,8:762227,doi:10.3389/frobt.2021.762227.

[12] Xie G, Yang L T, Liu Y, et al. Security enhancement for real-time independent in-vehicle CAN-FD messages in vehicular networks「J]. IEEE Transactions on Vehicular Technology, 2021, 70(6): 5244-5253.

[13] Ma W, Liu Y, Xie G, et al. Security-aware CAN-FD message pack- ing in intelligent automotive cyber-physical systems[J]. IEEE In- ternet of Things Journal, 2021, 9(22):22343-22356.

[14] Stumpf F, Pohl C, Hoettges D, et al. Introducing HSM-based secure on-board communication in vehicles[C]// Proceedings of Escar Europe, 2019: 104107.

[15] Xie Y, Guo Y, Yang S, et al. Security-related hardware cost optimi- zation for CAN FD-based automotive cyber-physical systems[J]. Sensors, 2021, 21 (20):6807,doi:10.3390/s21206807.

[16] Reusch N, Craciunas S S, Pop P. Dependability-aware routing and scheduling for time-sensitive networking[J]. IET Cyber-Physical Systems:Theory ＆ Applications, 2022, 7 (3):124-146.

[17] Dabbaghjamanesh M, Moeini A, Reinforcement learning-based load forecasting of electric vehicle charging station using Q-learning technique[J]. IEEE Transactions on Industrial In- formatics, 2021, 17(6):4229-4237.

[18] Xie G, Yang L T, Wu W, et al. Security enhancement for real-time parallel in-vehicle applications by CAN FD message authentication[J]. IEEE Transactions on Intelligent Transportation Systems, 2020, 22(8):5038-5049.

[19] LIU Y C. Research on multi-sensor data fusion optimization algo- rithm in the internet of things environment[J]. Technology of IoT ＆ AI, 2025, 57(2):33-36.

[20] Topcuoglu H, Hariri S, Wu M Y. Performance-effective and low- complexity task scheduling for heterogeneous computing[J]. IEEE Transactions on Parallel and Distributed Systems, 2002, 13 (3): 260-274.

[21] LI Y H, LIU P P, WANG W M, et al. Deep reinforcement learning task offloading method in edge end collaboration scenario[J]. Journal of Chinese Computer Systems, 2025, 46(2):280-288.

[22] Dearden R, Friedman N, Russell S. Bayesian Q-learning[J]. Asso- ciation for the Advancement of Artificial Intelligence, 1998:761-768,doi: 10.1007/RF00993412.

[23] Buttazzo G C. Hard real-time computing systems:predictable scheduling algorithms and applications(3rd ed.)[M]. Berlin: Springer, 2011.

[24] Chauhan N K, Tyagi I, Kumar H, el al. Tasks scheduling through hybrid genetic algorithm in real-time system on heterogeneous envi- ronment[J]. SN Computer Science, 2022, 3(1):75,doi: 10.1007/s42979-021-00959-0.

[25] Xie G, Zeng G, Kurachi R, et al. Exact WCRT analysis for mes- sage-processing tasks on gateway-integrated in-vehicle CAN clus- ters[J]. ACM Transactions on Embedded Computing Systems, 2018, 17(6):1-29.

[26] 刘裕灿. 物联网环境下多传感器数据融合优化算法研究[J]. 智能物联技术, 2025, 57(2):33-36.

[27] 李英豪, 刘盼盼, 王文猛, 等. 边端协同场景下的深度强化学习任务卸载方法[J]. 小型微型讣算机系统, 2025, 46(2):280-288.

基金资助

国家白然科学基金项目(62072175)