芯片间时间触发消息堆叠调度方法  

Stacking scheduling method for time-triggered messages in off-chip network

在线阅读下载全文

作  者:臧光界 李峭[2] 王彤[2] 熊华钢[2] ZANG Guangjie;LI Qiao;WANG Tong;XIONG Huagang(Chinese Aeronautical Establishment,Beijing 100029,China;School of Electronic Information Engineering,Beihang University,Beijing 100191,China)

机构地区:[1]中国航空研究院,北京100029 [2]北京航空航天大学电子信息工程学院,北京100191

出  处:《北京航空航天大学学报》2023年第7期1838-1846,共9页Journal of Beijing University of Aeronautics and Astronautics

基  金:国家自然科学基金(62071023)。

摘  要:时间触发(TT)通信方式应用于芯片间互连网络,以保证航空电子通信任务之间消息传递的严格时间确定性。当航空电子任务具有多种操作模式,属于不同模式的芯片间的时间触发调度表会重叠占用时隙,提出芯片间时间触发消息堆叠调度方法,以提高利用网络资源的灵活性和效率,同时减小应用层消息由于等待时间触发时间窗的排队延迟。仿真实验表明:与超调度方法相比,所提方法能够减小芯片间互连网络中时间触发消息的总端到端延迟和链路平均时隙占用率,对于端到端延迟时间较长且链路平均承载消息传输较多的场景,采用所提方法减少端到端延迟的效果更显著。To ensure the strict time determinism of message transmission between avionics communication tasks,time-triggered(TT)communication methods are also applied to the off-chip interconnection network.When avionics tasks have multiple operation modes,the time-triggered schedules between chips belonging to different modes will overlap and occupy time slots.The stacking scheduling method for time-triggered messages in off-chip network,which can reduce the queuing delay of application layer messages due to the waiting TT time window,is proposed to improve the flexibility and efficiency of using network resources,.Simulation experiments show that compared with the super scheduling method,the stack scheduling method can reduce the total end-to-end delay of TT messages in the off-chip interconnection network and the average time slot occupancy rate of the link.For scenarios where the end-to-end delay time is long and the average link bearer message transmission is large,the effect of using the stack scheduling method to reduce the end-to-end delay is more significant.

关 键 词:芯片间互连 时间触发消息 操作模式 堆叠调度 端到端延迟 

分 类 号:TP393[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象