Ethernet Network Fault Containment via Redundant Message Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Ethernet networks in safety-critical applications face challenges in ensuring operability and error tolerance, particularly in scenarios where components like computing nodes or star couplers become faulty, necessitating a solution that maintains network functionality and security even in fault conditions.
Innovation Solution
The implementation of a method where receiving components in an Ethernet network accept and forward Ethernet messages only if identical messages are received via multiple communication lines, utilizing redundant fault-containment units to ensure fault tolerance and error containment, with specific configurations for MAC and PHY layers, and employing standards like IEEE 1588 for time-controlled communication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundant communication lines and fault-containment units are implemented to improve error tolerance, then network reliability is improved, but device complexity increases
Solution Approach 1:
The network components are segmented into fault-containment units with distinct MAC and PHY layers. Each fault-containment unit operates independently, allowing faults to be contained within specific segments without affecting the entire network. This segmentation enables redundancy implementation while managing complexity through modular isolation.
Solution Approach 2:
The patent introduces redundancy in the communication dimension by implementing multiple communication lines between components. Messages are transmitted through multiple paths simultaneously, and receiving components compare messages from different lines to detect and correct errors, adding a dimensional layer of verification without fundamentally changing the network topology.
2Measurement precision
If multiple identical messages are required to be received via different communication lines to ensure accuracy, then message accuracy is improved, but productivity decreases
Solution Approach 1:
The system performs preliminary verification by requiring multiple identical messages to be received via different communication lines before accepting a message as valid. This preliminary action of cross-verification ensures message accuracy by detecting transmission errors, corruption, or interference before the message is processed further in the network.
Solution Approach 2:
The patent implements a copying mechanism where messages are transmitted through multiple redundant communication lines simultaneously. The receiving component obtains copies of the same message through different paths and compares them for identity. This copying approach ensures accuracy without requiring retransmission, as all copies are sent in parallel.
3Reliability
If fault-containment units with separate MAC and PHY layers are implemented to improve error containment, then error tolerance is improved, but device complexity increases
Solution Approach 1:
Each fault-containment unit is segmented into distinct MAC (Media Access Control) and PHY (Physical) layers, allowing independent operation and fault isolation. If a fault occurs in the PHY layer, it does not affect the MAC layer or other fault-containment units. This segmentation enables precise error containment while managing complexity through functional separation.
Solution Approach 2:
The fault-containment units act as intermediaries between the physical communication medium and the higher-level network protocols. By introducing this intermediate layer with separate MAC and PHY components, the system can isolate and contain errors at the fault-containment unit level, preventing them from propagating through the entire network while maintaining standard Ethernet protocol compatibility.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
The invention relates to a method for transmitting messages in a computer network and to a computer network of this kind. The computer network comprises computation nodes (101-105), which computation nodes (101-105) are connected to one another via at least one star coupler (201) and/or at least one multi-hop network (1000), each computation node (101-105) being connected to the at least one star coupler (201) and/or to the at least one multi-hop network (1000) via at least one communication line (110), and the computation nodes (101-105) exchanging Ethernet messages with one another and with the at least one star coupler (201) and/or the at least one multi-hop network (1000). Provision is made for a set of two or more components each to be directly connected to one another by two or more communication lines (110, 111), each component in the set being either a computation node (101-105) or a star coupler (201), and sending components in the set of components sending at least some of the Ethernet messages that are to be sent to at least two of the two or more communication lines (110, 111), and receiving components in the set of components accepting and/or forwarding at least some of the Ethernet messages that are received via the two or more communication lines (110, 111) only if at least two identical messages are received via at least two different communication lines.