Quality-of-Service mechanisms are required for the transmission of flows with different criticality levels on a real-time Ethernet network. Static Priority Queueing (SPQ) is such a popular mechanism since it can bound the impact of lower-criticality flows on higher-criticality ones. To ensure deadlines for critical flows, several worst-case traversal time (WCTT) analysis methods have been proposed for SPQ. However, they compute pessimistic upper bounds on delays, causing unnecessary network over-dimensioning. This paper identifies and analyses the sources of pessimism in Network Calculus based WCTT analysis under SPQ. We also propose a method to construct near-worst-case delay scenarios within a tractable search space, enabling estimation of an upper bound on pessimism. Our results reveal a significant gap between computed upper bounds and attainable near-worst-case delays, paving the way for improvements in WCTT analysis in real-time Ethernet networks.

Real-time Ethernet has become a popular solution for handling critical communication in embedded systems, partly thanks to the availability of safe worst-case traversal time analysis. These analyses scale well, but provide pessimistic upper bounds on end-to-end delays leading to over-dimensioning of the network architecture. Computing worst-case delays requires an exhaustive search over all possible scenarios that quickly leads to a combinatorial explosion. Consequently, it is limited to network configuration with less than 100 flows, far from industrial-size ones with more than 1000 flows. In this paper, we propose a hybrid approach (HA) that (1) reduces the number of scenarios to be analysed to get the worst-case delay, (2) leverages exact delays in order to tighten bounds on worst-case delays when the reduced number of scenarios is still too big to be completely analysed. On a realistic avionics case study, HA proved scalable for worst-case delay computations and reduced upper-bound pessimism by over 40%.

Deficit Round-Robin (DRR) is a promising service discipline for real-time Ethernet without a global synchronisation. Two improved Network Calculus approaches have been proposed to provide the required bounds on end-to-end delays. The first one is fast but can be optimistic for cornet cases. The second one is safe but highly time consuming. In this paper, we remove the potential optimism of the first approach while keeping its low complexity.

In a real-time network shared between flows with different criticality, the service discipline in network elements highly impacts the overall performance of the system. Weighted Round Robin (WRR) is a well known simple scheduling policy. It has much lower complexity and cost than Deficit Round Robin (DRR) and exhibits GPS like fairness in a fixed packet size environment.In this paper, we consider WRR in an avionics context where data flows of critical avionic functions must co-exist with noncritical flows in a switched Ethernet network, called AFDX. These flows can have heterogeneous delay constraints and features. Upper bounding the worst-case delays of avionic flows is a mandatory part of the certification process. Meanwhile, sharing AFDX with non-critical flows is envisioned to improve bandwidth utilisation. We show that configuring WRR in such environment is challenging, since frame sizes and deadlines can be very different. We propose a heuristic to distribute critical flows in classes and an exhaustive approach to allocate weights to these classes. Our goal is to achieve maximum bandwidth for noncritical flows while respecting the critical flow deadlines.

AFDX is the de facto communication standard in avionics domain. It is primarily used for the transmission of critical avionic flows. The mandatory certification process requires to upper bound the end-to-end transmission delay for each critical flow. Therefore, worst-case traversal time analysis has been implemented. However, it leads to a very lightly loaded network (up to 10%) as it considers very rare worst-case situations. Introducing a QoS mechanism is often a good solution to improve network utilisation since it allows differentiating critical flows based on their constraints as well as the transmission of less/non-critical flows with bounded impact on critical ones. Deficit Round Robin is such a mechanism and it is envisioned for future avionic networks. Using this mechanism, we propose to share the bandwidth between n - 1 classes for critical flows and one class for non-critical ones. Therefore the contributions of this paper are (1) to propose a quantum assignment that ensures that critical flows always respect their deadlines and maximises the bandwidth for non-critical ones and (2) to propose a heuristic for the distribution of the flows among different classes.

In order to handle mixed criticality flows in a realtime embedded network, switched Ethernet with Quality of Service (QoS) facilities has become a popular solution. Deficit Round Robin (DRR) is such a QoS facility. Worst-Case Traversal Time (WCTT) analysis is mandatory for such systems, in order to ensure that end-to-end delay constraints are met. Network Calculus is a classical approach to achieve this WCTT analysis. A solution has been proposed for switched Ethernet with DRR. It computes pessimistic upper bounds on end-to-end latencies. This pessimism is partly due to the fact that the scheduling of flows by end systems is not considered in the analysis. This scheduling can be modeled by offsets between flows. This modeling has been integrated in WCTT analysis of switched Ethernet with First In First Out (FIFO) scheduling. It leads to a significant reduction of delay upper bounds. The contribution of this paper is to integrate the offsets in the WCTT analysis for switched Ethernet with DRR and to evaluate the reduction on delay upper bounds, considering a realistic case study.

Avionics Full Duplex switched Ethernet (AFDX) is the de facto standard for the transmission of critical avionics flows. It is a specific switched Ethernet solution based on First-in First-out (FIFO) scheduling. Worst-case traversal time (WCTT) analysis is mandatory for such flows, since timing constraints have to be guaranteed. A classical approach in this context is Network Calculus (NC). However, NC introduces some pessimism in the WCTT computation. Moreover, the worst-case often corresponds to very rare scenarios. Thus, the network architecture is most of the time lightly loaded. Typically, less than 10 % of the available bandwidth is used for the transmission of avionics lows on an AFDX network embedded in an aircraft. One solution to improve the utilization of the network is to introduce Quality of Service (QoS) mechanisms. Deficit Round Robin (DRR) is such a mechanism and it is envisioned for future avionics networks. A WCTT analysis has been proposed for DRR. It is based on NC. It doesn't make any assumption on the scheduling of flows by end systems. The first contribution of this paper is to identify sources of pessimism of this approach and to propose an improved solution which removes part of this pessimism. The second contribution is to show how the scheduling of flows can be integrated in this optimized DRR approach, thanks to offsets. An evaluation on a realistic case study shows that both contributions bring significantly tighter bounds on worst-case latencies.