Using Holistic Queue Scheduling to Optimize the Real-Time Performance of AFDX

AFDX, Avionics Full Duplex Switched Ethernet Network, is an aircraft data network using the FIFO scheduling strategy to transmit data. In this circumstance, the delay of the critical data is large, while congestion is uncertain. To cope with this problem, this paper presents a Holistic Queue Scheduling policy to optimize the real-time performance of AFDX, and estimate the latency bounds based on Network Calculus. Simulation results show that this policy not only ensures real-time critical data, but also ensures the fairness of scheduling non-critical data. And this policy can reduce the impact of malicious data in some extent.