Performance Enhancement and Optimized Analysis of the Worst Case End-to-End Delay for AFDX Networks

Abstract

Avionics Full Duplex Switched Ethernet (AFDX) network is an emerging Ethernet media based technology for the next generation avionic data bus. One of the key issues in AFDX networks is to guarantee the data transmission delay and optimize the worst case end-to-end delay performance in order to meet the requirements of safety-critical applications, such as flight control. In this paper, a more accurate Virtual Link (VL) model, the Generic Cell Rate Algorithm (GCRA), is used to derive a tighter worst case end-to-end delay bound analysis. A hierarchical scheduling policy of Fixed Priority combined with Round Robin (FP-RR) is then applied to AFDX network entities, and the corresponding end-to-end delay bound for different priority classes is derived. Finally, the proposed approaches are validated on a True Time-based simulator platform. The results show that the new GCRA model leads to improved performance, and the new end-to-end delay bound analysis under FP-RR scheduling policy has higher accuracy in network performance evaluation for separate priority classes.