航空学报 > 2015, Vol. 36 Issue (6): 1924-1932   doi: 10.7527/S1000-6893.2014.0323

AFDX虚拟链路路径实时寻优算法

代真, 何锋, 张宇静, 熊华钢   

  1. 北京航空航天大学 电子信息工程学院, 北京 100191
  • 收稿日期:2014-07-16 修回日期:2014-09-23 出版日期:2015-06-15 发布日期:2014-11-29
  • 通讯作者: 何锋 Tel.: 010-82338894 E-mail: robinleo@buaa.edu.cn E-mail:robinleo@buaa.edu.cn
  • 作者简介:代真 男, 硕士研究生。主要研究方向: 机载网络。 Tel: 010-82338712 E-mail: daizhen@buaa.edu.cn;何锋 男, 博士, 讲师。主要研究方向: 航空电子综合, 实时调度、实时网络。 Tel: 010-82338894 E-mail: robinleo@buaa.edu.cn;张宇静 女, 硕士研究生。主要研究方向: 机载网络。 E-mail: zhangyujing327@gmail.com;熊华钢 男, 博士, 教授, 博士生导师。主要研究方向: 航空电子综合、机载网络。 Tel: 010-82317202 E-mail: hgxiong@ee.buaa.edu.cn
  • 基金资助:

    国家自然科学基金 (61301086); 航空科学基金 (20131951027)

Real-time path optimization algorithm of AFDX virtual link

DAI Zhen, HE Feng, ZHANG Yujing, XIONG Huagang   

  1. School of Electronics and Information Engineering, Beihang University, Beijing 100191, China
  • Received:2014-07-16 Revised:2014-09-23 Online:2015-06-15 Published:2014-11-29
  • Supported by:

    National Natural Science Foundation of China (61301086); Aeronautical Science Foundation of China (20131951027)

摘要:

航空电子全双工交换式以太网(AFDX)使用虚拟链路(VL)进行消息数据流通信,为提高AFDX的网络实时传输性能,以VL路径配置寻优为基础,提出了一种基于遗传算法的AFDX VL路径优化算法(POGA)。POGA以提高网络实时性为优化目标,并综合考虑网络负载均衡,以VL的路径信息作为染色体,通过遗传算子进行遗传寻优操作,选择促使AFDX网络实时性能最优的VL路径作为优化结果。利用仿真优化方法对POGA进行了实现,在典型AFDX网络1 000条VL的配置下,与负载均衡算法和最短路径算法进行了分析对比,结果显示分别有76.4% 和77.4%的VL的传输实时性得到了增强,网络的实时传输性能分别提高了13.2%和12.9%。

关键词: 航空电子全双工交换式以太网, 虚拟链路, 路径配置优化, 遗传算法, 网络性能

Abstract:

Avionics full duplex switched ethernet (AFDX) uses virtual link (VL) for message communications. In order to improve the real-time transmission performance of AFDX networks, with the target of VL path configuration optimization, a virtual link path optimization algorithm based on genetic algorithm (POGA) is proposed. In POGA, the network's real-time transmission performance is used as optimization objectives, and the network load balancing has also been considered. With VL path used as chromosome, POGA works through genetic operation and tries to find out the optimal VL path solution which makes the AFDX's real-time network performance the best. POGA is implemented by simulation. Under a configuration of 1 000 VLs of a typical AFDX network, POGA is compared with the load balancing algorithm and the shortest path algorithm, and results show that 76.4% and 77.4% of VLs' real-time transmission latency has been reduced, and the real-time transmission performance of the AFDX network is enhanced by 13.2% and 12.9% respectively.

Key words: AFDX, virtual link, path configuration optimization, genetic algorithms, network performance

中图分类号: