Ethernet Switch Selective Buffering for Processor Startup
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Ethernet switches face challenges in managing data capacity during startup when processors with varying activation times are connected, potentially leading to data loss due to buffer overflow.
Innovation Solution
An Ethernet switch with a controller and buffers that selectively holds and transmits data based on predetermined conditions, ensuring only target data is stored until communication is established with slower processors, and discarding non-target data to prevent overflow.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the Ethernet switch holds all received data in the buffer during processor activation, then data transmission reliability is improved, but buffer overflow occurs and important data cannot be held
Solution Approach 1:
The patent extracts and identifies target data based on destination processor information before buffering. The controller determines which data should be held in the buffer by extracting relevant identification information (such as destination processor ID) from the received data, allowing selective buffering of only necessary data rather than all data, thus preventing buffer overflow while maintaining transmission reliability
Solution Approach 2:
The patent applies different handling qualities to different data based on their destination. The controller examines the destination processor identification information and applies selective buffering policies: data destined for processors that will be activated soon is buffered, while data for already-active processors is immediately transmitted. This local differentiation of data handling quality optimizes buffer usage while ensuring reliable transmission
2Loss of information
If the Ethernet switch buffers all received data during processor activation, then data loss is prevented, but the buffer capacity is exceeded and non-critical data occupies space
Solution Approach 1:
The controller extracts destination processor identification information from received data packets and uses this extracted information to determine which packets should be buffered. By extracting and analyzing the destination identifier, the system can selectively buffer only the necessary data for processors that need it, preventing both data loss and unnecessary buffer occupancy
Solution Approach 2:
The system performs preliminary determination of target data before buffering occurs. The controller preliminarily identifies which data should be buffered by examining destination processor status and activation timing information, allowing the buffer to be filled only with appropriate data in advance, thus preventing buffer overflow while ensuring critical data is not lost
Data Source
AI summary
An Ethernet switch connected to a plurality of on-board electronic control units each having at least one processor and having a plurality of ports connecting the respective processors and configured to receive data from each processor and transmit data to a destination processor comprises a controller configured to determine target data from the received data based on a predetermined condition, and a buffer configured to hold the target data. The controller is configured such that, when the target data is determined, the target data is held in the buffer until communication with the destination processor is permitted, when communication with the destination processor is permitted, the target data is transmitted to the destination processor, and when data different from the target data is received before communication with the destination processor is permitted, the received data is not held in the buffer.