Integrated planning method and optimization for off-chip time-triggered communication
-
摘要:
随着片上系统(SoC)的处理能力逐渐接近传统的综合核心处理模块,航空电子系统向着微小型综合化的芯片间系统发展;时间触发交换式互连可以保证芯片间消息传递的严格时间确定性。考虑芯片间互连交换结构轻量化和收发端口有限的特点,在拓扑、路由和调度时刻等网络资源相互制约的条件下,提出了芯片间时间触发通信综合规划方法,即根据时间触发消息集合和芯片端口配置,同时求解得到芯片间网络拓扑结构、消息路由和调度时刻表的规划结果。其中,采用免疫算法整体优化了各条消息在网络资源分配过程中的求解次序。仿真实验表明,与不考虑整体优化的综合规划方法相比,优化后的规划结果在减少拓扑结构中多余路径开销的同时,避免消息传输路径拥堵,降低消息端到端延迟,保证了消息集的可调度性。
Abstract:As the processing capacity of system-on-chip (SoC) is getting close to the traditional integrated core processing module, the avionics system is developing towards the miniature, integrated off-chip system. Time-triggered switched interconnection can guarantee the strictly time deterministic property of off-chip message transmission. Considering the off-chip interconnection characteristics of lightweight switch structure and limited port number on a chip, an integrated planning method for off-chip time-triggered communication was proposed under the condition of interaction among topology planning, routing and scheduling. Given time-triggered message sets and port configuration, the off-chip interconnection network topology, message routing and scheduling table were obtained at the same time. Then the message allocation order is optimized using immune algorithm to further improve the performance of the algorithm. The simulation results show that, compared with the integrated planning method without consideration of overall optimization, the optimized method reduced the congestion on message transmission paths, reduced the message end-to-end delay and increased the schedulability of message sets while generating the off-chip interconnection topology with low costs.
-
表 1 消息集合示例
Table 1. Example of message set
消息编号 周期/μs 长度/Byte 源节点 目的节点 1 400 23 1 4 2 400 683 2 4 3 100 433 3 4 4 400 1 429 1 2 5 100 183 2 4 6 400 233 1 5 7 200 483 2 6 8 50 43 2 1 9 100 124 5 3 10 400 345 6 2 表 2 实验组别参数
Table 2. Parameters of experimental groups
消息组别 芯片个数 消息个数 消息长度范围/ Byte 周期范围/μs 1 6 10~40 23~1 483 50~400 2 8 10~40 14~1 468 50~400 3 10 10~40 13~1 454 50~400 4 12 10~40 11~1 465 50~400 5 14 10~40 35~1 457 50~400 表 3 优化TRS综合规划方法和增量化SMT方法可调度性对比
Table 3. Schedulability comparison between optimized TRS and incremental SMT
消息组别 增量化SMT方法消息个数 优化TRS综合规划方法消息个数 10 20 30 40 10 20 30 40 消息组别1(6芯片个数) √ × × × √ √ √ √ 消息组别2(8芯片个数) √ √ √ × √ √ √ √ 消息组别3(10芯片个数) √ √ √ × √ √ √ √ 消息组别4(12芯片个数) √ √ × × √ √ √ √ 消息组别5(14芯片个数) √ √ × × √ √ √ √ -
[1] 熊华钢, 王中华.先进航空电子综合技术[M].北京:国防工业出版社, 2009:2-13.XIONG H G, WANG Z H.Advanced avionics integration techniques[M].Beijing:National Defense Industry Press, 2009:2-13(in Chinese). [2] WOLFIG R, JAKOVLJEVIC M.Distributed IMA and DO-297: Architectural, communication and certification attributes[C]//Proceedings Digital Avionics Systems Conference.Piscataway, NJ: IEEE Press, 2008: 1.E.4-1-1.E.4-10. [3] 蒲小勃.现代航空电子系统与综合[M].北京:航空工业出版社, 2013:70-86.PU X B.Modern avionics system and integration[M].Beijing:Aviation Industry Press, 2013:70-86(in Chinese). [4] YEOMANS J, TROTTET A.E2v and Adeneo partner to create the world's smallest, multicore computer for aerospace applications[EB/OL].(2016-07-18)[2019-03-27].https://www.e2v.com/news/e2v-and-adeneo-partner-to-create-the-worlds-smallest-multicore-computer-for-aerospace-applications/. [5] ASHLEY.Avionics systems hosted on a distributed modular electronics large scale demonstrator for multiple types of aircraft[EB/OL].(2015-11-05)[2019-03-27].http://www.ashleyproject.eu/ashley-event-2015/ebook/index.php. [6] ABUTEIR M, OBERMAISSER R, OWDA Z, et al.Off-chip/on-chip gateway architecture for mixed-criticality systems based on networked multi-core chips[C]//International Conference on Computational Science and Engineering.Piscataway, NJ: IEEE Press, 2015: 120-128. [7] OLIVER R S, CRACIUNAS S S.Hierarchical scheduling over off-and on-chip deterministic networks[J].ACM SIGBED Review, 2016, 13(4):14-19. doi: 10.1145/3015037.3015039 [8] 何锋.机载网络技术基础[M].北京:国防工业出版社, 2018:192-210.HE F.Fundamentals of airborne network[M].Beijing:National Defense Industry Press, 2018:192-210(in Chinese). [9] 杨飞生, 汪璟, 潘泉.基于事件触发机制的网络控制研究综述[J].控制与决策, 2018, 33(6):969-977. http://d.old.wanfangdata.com.cn/Periodical/kzyjc201806001YANG F S, WANG J, PAN Q.A survey of networked event-triggered control[J].Control and Decision, 2018, 33(6):969-977(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/kzyjc201806001 [10] 杨飞生, 汪璟, 潘泉, 等.网络攻击下信息物理融合电力系统的弹性事件触发控制[J].自动化学报, 2019, 45(1):110-119. http://d.old.wanfangdata.com.cn/Periodical/zdhxb201901009YANG F S, WANG J, PAN Q, et al.Resilient event-triggered control of grid cyber-physical systems against cyber attack[J].Acta Automatica Sinica, 2019, 45(1):110-119(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/zdhxb201901009 [11] STEINER W.An evaluation of SMT-based schedule synthesis for time-triggered multi-hop networks[C]//Real-Time Systems Symposium.Piscataway, NJ: IEEE Press, 2010: 375-384. [12] 宋梓旭, 李峭, 汪晶晶, 等.基于可调度性排序的时间触发调度表生成方法[J].北京航空航天大学学报, 2018, 44(11):2388-2395. doi: 10.13700/j.bh.1001-5965.2018.0043SONG Z X, LI Q, WANG J J, et al.Time-triggered scheduling table generation method based on schedulability ranking[J].Journal of Beijing University of Aeronautics and Astronautics, 2018, 44(11):2388-2395(in Chinese). doi: 10.13700/j.bh.1001-5965.2018.0043 [13] 徐晓飞, 曹晨, 郭骏, 等.TT-RMS:时间触发网络通信表生成算法[J].北京航空航天大学学报, 2015, 41(8):1403-1408. doi: 10.13700/j.bh.1001-5965.2014.0618XU X F, CAO C, GUO J, et al.TT-RMS:Communication table generation algorithm of time-triggered network[J].Journal of Beijing University of Aeronautics and Astronautics, 2015, 41(8):1403-1408(in Chinese). doi: 10.13700/j.bh.1001-5965.2014.0618 [14] TAMAS-SELICEAN D, POP P, STEINER W.Design optimization of TTEthernet-based distributed real-time systems[J].Real-Time Systems, 2015, 51(1):1-35. doi: 10.1007/s11241-014-9214-8 [15] SMIRNOV F, GLAB M, REIMANN F, et al.Optimizing message routing and scheduling in automotive mixed-criticality time-triggered networks[C]//Design Automation Conference.Piscataway, NJ: IEEE Press, 2017: 1-6. [16] POP P, RAAGAARD M L, CRACUINAS S S, et al.Design optimization of cyber-physical distributed systems using IEEE time-sensitive networks[J].IET Cyber-Physical Systems:Theory & Applications, 2016, 1(1):86-94. http://cn.bing.com/academic/profile?id=a44751f901c2586e59a5c9849bac0292&encoded=0&v=paper_preview&mkt=zh-cn [17] GAVRILUT V, ZARRIN B, POP P, et al.Fault-tolerant topology and routing synthesis for IEEE time-sensitive networking[C]//Proceedings of the 25th International Conference on Real-Time Networks and Systems.New York: ACM, 2017: 267-276. [18] ATALLAH A A, HAMAD G B, MOHAMED O A.Fault-resilient topology planning and traffic configuration for IEEE 802.1Qbv TSN networks[C]//International Symposium on On-Line Testing and Robust System Design.Piscataway, NJ: IEEE Press, 2018: 151-156. [19] 孔韵雯, 李峭, 熊华钢, 等.片间综合化互连时间触发通信调度方法[J].航空学报, 2018, 39(2):321590. http://d.old.wanfangdata.com.cn/Periodical/hkxb201802023KONG Y W, LI Q, XIONG H G, et al.Time-triggered communication scheduling method for off-chip integrated interconnection[J].Acta Aeronautica et Astronautica Sinica, 2018, 39(2):321590(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/hkxb201802023 [20] 包子阳, 余继周, 杨杉.智能优化算法及其MATLAB实例[M].2版.北京:电子工业出版社, 2018:60-85.BAO Z Y, YU J Z, YANG S.Intelligent optimization algorithm and its MATLAB example[M].2nd ed.Beijing:Publishing House of Electronics Industry, 2018:60-85(in Chinese). [21] YEN J Y.An algorithm for finding shortest routes from all source nodes to a given destination in general networks[J].Quarterly of Applied Mathematics, 1970, 27(4):526-530. doi: 10.1090/qam/253822 [22] YEN J Y.Finding the K shortest loopless paths in a network[J].Management Science, 1971, 17(11):712-716. doi: 10.1287/mnsc.17.11.712 [23] JERNE N K.Towards a network theory of the immune system[J].Annual Immunology, 1974, 125:373-389. [24] EVEN S, GOLDREICH O, MORAN S, et al.On the NP-completeness of certain network testing problems[J].Networks, 1984, 14(1):1-24 doi: 10.1002/net.3230140102