期刊文献+

基于PCIeDMA缓冲池的流量控制协议

Flow Control Protocol Based on PCIe DMA Buffer Pool
下载PDF
导出
摘要 随着航空电子的发展,机载计算机对通信性能提出了更高的要求。PCIe总线的直接存储器存取(Direct Memory Access,DMA)通信方式凭借高吞吐量、低时延及低中央处理器(Central Processing Unit,CPU)占用率等优势,广泛应用于嵌入式计算机通信。在复杂网络拓扑的PCIe架构中,由于通信节点接收数据和发送数据的协议差异或中间介质差异,导致入向流量和出向流量速率不匹配,造成数据缺失。因此,文章设计了一种基于PCIeDMA缓冲池的流量控制协议,以控制模块间的通信流量。利用多级缓冲技术,在节点中配置环形缓冲池,可回收利用总线空间。利用同步互斥技术为高速数据转发争取时间,避免了多节点设备转发数据出现滞后性和数据缺失问题。 With the development of avionics,airborne computers put forward higher requirements for communication performance.The Direct Memory Access(DMA)communication mode of PCIe bus is widely used in embedded computer communication with the advantages of high throughput,low latency and low Central Processing Unit(CPU)utilization.In the PCIe architecture with complex network topology,due to the protocol difference or medium difference between the communication nodes receiving and sending data,the incoming traffic and outgoing traffic rates do not match,resulting in data loss.Therefore,this paper designs a flow control protocol based on PCIe DMA buffer pool to control the communication flow between modules.The bus space can be recycled by using multi-level buffer technology and configuring ring buffer pool in nodes.Using synchronous mutual exclusion technology to gain time for high-speed data forwarding,to avoid the problems of lag and data loss in multi-node equipment forwarding data.
作者 黄双双 郝一太 罗伟杰 HUANG Shuangshuang;HAO Yitai;LUO Weijie(Aeronautics Computing Technology Research Institute Xi’an,Xi’an 710068,China)
出处 《通信电源技术》 2024年第3期16-18,共3页 Telecom Power Technology
关键词 PCIE 直接存储器存取(DMA) 多级缓冲 同步互斥 PCIe Direct Memory Access(DMA) multi-level buffer synchronous mutual exclusion
  • 相关文献

参考文献7

二级参考文献36

  • 1Syed ljlal Ali Shah, Shashank Khanvilkar, Ashfaq Khokhar. RapidI() traffic management and flow arbi- tration protocol EJ]. IEEE Communications Maga- zine, 2006,44(7) :45-51.
  • 2Sam Fuller. RapidI(): The embedded system interconnect [M]. UK: John Wiley ~-. Sons, 2005.
  • 3The RapidlO trade association. RapidlO interconnect specifi cation part6[S]. Rev 2. 1. 2009.
  • 4Xilinx I.ogiC()RE IP .serial rapidIO (~a2 endpoint v2. 0. product guide for vivado Design Suite[R~. 2013.
  • 5X B Zhao, S Jia, Y Wang, et al. Buffer design based on flow control in rapidI(~C~//2010 Asia Pacific conference on Postgradvate Research in Microelectronics and Elec- t ronics Shanghai: IEEE, 2010 : 150-153.
  • 6David Bueno, Adam Leko, Chris Conger, et al. Simulative analysis of the rapidI() embedded interconnect architecture for real-time, network-imerksive applicatiott~ [C]///Pro- ceedings of the 29th Annual IEEE International Conference on Iaxal Computer Networks Tampa, Fix, USA.. IEL~, 2004.710-727.
  • 7Wu Fengfeng, Jia Song, Wang Yuaix An effective buffer space management in serial rapidI()endpoint[C~//2011 A- sia Pacific Conference on Postgraduate Research in Microe- lectronics and Electronics Maeaa: IEE'E, 2011 : 135-138.
  • 8白秀玲,宋晓莉,邱涌,张蕾,张孝国,杨春蕾.进程同步的教学方法探讨[J].计算机时代,2009(9):71-73. 被引量:2
  • 9高常波.Vxworks下多串口扩展卡驱动实现[J].信息安全与通信保密,2010,7(4):47-49. 被引量:2
  • 10强新建,田泽,淮治华.VxWorks下驱动程序设计[J].西安石油大学学报(自然科学版),2010,25(3):82-84. 被引量:2

共引文献29

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部