http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Qingfei Xu,Xinyu Yang 한국항공우주학회 2019 International Journal of Aeronautical and Space Sc Vol.20 No.2
With growing avionics applications, information transmission has been increasing in real-time systems of airplanes; AFDX (Avionics Full Duplex Switched Ethernet) standardized as ARINC 664 is widely used in avionics transmission systems for its high speed and duplex characteristics. For safety–critical systems such as flight control system, the time limit of information transmission should be deterministic and it is used to evaluate the performance of the real-time systems. Network Calculus is a powerful approach used to estimate the performance of avionics real-time system by calculating the possible longest transmission time, but the performance analysis is usually pessimistic for the pessimistic assumptions in the computation. Grouping strategy can improve that analysis by considering serialization effect of information transmitting from the same physical link, but the analysis of system performance is still pessimistic. In this paper, we explore the pessimism in Network Calculus and Grouping strategy, and then propose a Rate-constrained grouping strategy to improve the analysis of system performance. The experiments on both sample avionics system and real industrial system confirm the validity and applicability of the performance analysis.
Qingfei Xu,Xinyu Yang 한국항공우주학회 2020 International Journal of Aeronautical and Space Sc Vol.21 No.4
AFDX (Avionics Full Duplex switched Ethernet) standardized as ARINC 664 is chosen as the backbone network for distributed real-time avionics systems as it offers high throughput and does not require global clock synchronization. A deterministic upper bound of the end-to-end transmission delays for packets of each flow should be guaranteed to ensure the network performance. In this paper, we focus on the forward end-to-end delay analysis (FA). This approach iteratively estimates the maximum backlog (amount of the pending packets) in the switch output port along the transmission path so that the worst-case end-to-end transmission delay can be computed and the network performance can be evaluated. Recent research demonstrates this approach is pessimistic (overestimated). This paper presents an optimization which considers the serialization effect of packets transmitted through the same physical link. We provide a comparative analysis on both sample network configurations and real industrial systems. The experiments show that by considering the serialization effect the forward end-to-end delay analysis can remove the pessimism in the analysis of end-to-end transmission delay. The evaluation of network performance can be improved. Then a discussion on the potential optimism (underestimation) is presented.