Heise, Peter: Real-time guarantees, dependability and self-configuration in future avionic networksEchtzeit-Garantien, Zuverlässigkeit und Selbstkonfiguration in zukünftigen Avionik-Netzwerken. 2018
Inhalt
- Abstract
- Zusammenfassung
- Contents
- 1 Introduction
- 2 Basic Concepts
- 2.1 Dependability
- 2.2 Certification
- 2.2.1 Development Assurance Levels
- 2.2.2 Safety Analysis
- 2.2.3 Mixed Criticality Certification
- 2.2.4 Commercial off-the-shelf Devices
- 2.3 Real-Time Communication Systems
- 2.3.1 Mechanisms for Bounded Latency
- 2.3.1.1 Collision Detection
- 2.3.1.2 Time-Based Mechanisms
- 2.3.1.3 Token-Based & Master-Slave Mechanisms
- 2.3.1.4 Switched Mechanisms
- 2.3.2 Heterogeneous Latency Requirements
- 2.4 Verification of Performance Bounds
- 2.5 Summary
- 3 Related Work
- 3.1 Requirements and Challenges
- 3.2 Aeronautical Networks
- 3.2.1 Strictly Aeronautical Networks
- 3.2.1.1 Avionics Full-Duplex Switched Ethernet
- 3.2.1.2 Time Triggered Protocol
- 3.2.1.3 Legacy Networks
- 3.2.2 Adopted Automotive & Industrial Networks
- 3.2.3 Conclusion with Research Gap
- 3.3 Self-Configuration
- 4 Future Avionic Networks
- 4.1 Priority-Based Forwarding in Switches with a Global Timebase
- 4.1.1 Definitions
- 4.1.2 Mathematical Model
- 4.1.2.1 One TAQ per Node
- 4.1.2.2 Multiple TAQs per Node
- 4.1.2.3 Specifics to High Availability Seamless Redundancy
- 4.1.2.4 Gain of Time-Aware Queues
- 4.1.3 Simulation of Extended HSR
- 4.1.4 Comparison of Mathematical Model and Simulation Results
- 4.1.4.1 Normal Working Mode without Failure
- 4.1.4.2 Single Link Failure Case
- 4.1.4.3 Loss of Time Synchronization
- 4.1.5 Conclusion
- 4.2 Real-Time OpenFlow Network with Non-Synchronized Switches
- 4.2.1 Definition & Concept
- 4.2.1.1 Token Bucket in AFDX and OpenFlow
- 4.2.1.2 Deterministic Behavior and End-to-End Latencies
- 4.2.1.3 Matching of Virtual Links
- 4.2.1.4 Role of the OpenFlow Controller
- 4.2.2 Experiments and Results
- 4.2.2.1 Hardware vs. Software Matching
- 4.2.2.2 Meter Accuracy
- 4.2.2.3 Multiple Metered Streams
- 4.2.2.4 Failing End-System
- 4.2.3 Real-World Scenario
- 4.2.4 Conclusion
- 4.3 IEEE Time Sensitive Networking (TSN) with Fault-Tolerance and Frame Preemption
- 4.3.1 Overview of IEEE Time Sensitive Networking (TSN)
- 4.3.1.1 Explicit Stream Identification
- 4.3.1.2 Per-Stream Filtering and Policing (PSFP)
- 4.3.1.3 Frame Replication and Elimination for Reliability (FRER)
- 4.3.1.4 Frame Premption (Qbu)
- 4.3.2 TSimNet Simulation Framework
- 4.3.3 Evaluation
- 4.3.3.1 Simple Line Congestion
- 4.3.3.2 Industrial Line-Topology Preemption
- 4.3.3.3 Industrial TSN Profile
- 4.3.4 Conclusion
- 4.4 Results & Evaluation
- 5 Self-Configuring Network
- 5.1 System Model
- 5.2 Self-Configuring Plug and Play Real-Time Network
- 5.2.1 Concept & System Description
- 5.2.1.1 Network Control and Topology Awareness
- 5.2.1.2 Service Discovery
- 5.2.1.3 Constraining Traffic
- 5.2.1.4 Reaction to Failures
- 5.2.2 Implementation
- 5.2.3 Discussion & Evaluation
- 5.2.4 Conclusion
- 5.3 A Network with In-band Configuration
- 5.4 Results & Evaluation
- 6 Conclusion & Future Work
- A Author's Contribution
- B Bibliography