Frame Processing Device MAC Address Learning Order Reversal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In Ethernet frame transfer processing, the reversal of frame order between input and output occurs when learning is performed on un-learned frames, leading to inefficiencies as frame transfer speed increases, particularly due to the registration of MAC addresses in the MAC address table, causing un-learned frames to be processed before and after registration, resulting in order reversal and potential network bandwidth wastage.
Innovation Solution
A frame processing device that includes a learned frame passing information storage unit to record the destination MAC addresses of learned frames and check for their presence when un-learned frames are processed, discarding un-learned frames if their destination MAC address has already been learned, thus preventing order reversal and ensuring frames are transferred correctly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If flooding processing is performed on un-learned frames, then the destination device can receive the frame, but frame order reversal occurs when learning is performed during transfer
Solution Approach 1:
The patent applies preliminary action by checking whether the destination MAC address is already learned before performing flooding processing on un-learned frames. The learning determination unit checks the MAC address table in advance, and only frames with truly un-learned destination addresses are flooded, preventing order reversal caused by subsequent learning during transfer.
Solution Approach 2:
The patent uses feedback by continuously monitoring the MAC address table status during frame transfer. The learning determination unit receives feedback about whether destination addresses are learned, and dynamically adjusts processing - discarding frames if the address becomes learned during transfer, thereby maintaining frame order consistency.
2Stability of the object's composition
If un-learned frames are discarded when destination address is already learned, then frame order reversal is prevented, but additional processing overhead is introduced
Solution Approach 1:
The patent extracts the learning determination function as a separate unit that independently checks MAC address table status. This modular approach adds minimal processing overhead by performing simple table lookups rather than complex decisions, while effectively preventing frame order reversal.
Solution Approach 2:
The patent uses copying by creating a learned frame passing information storage unit that stores copies of learned MAC addresses. This allows rapid comparison and determination without complex processing, as the system simply checks whether destination addresses exist in the stored copy of learned addresses.
3Productivity
If learned frames are held in buffer while un-learned frames are processed, then frame transfer can be managed, but network bandwidth is wasted due to delayed learned frames
Solution Approach 1:
The patent applies dynamics by making frame transfer timing adaptive based on real-time learning status. Learned frames are transferred immediately when their destination address is learned, rather than being statically held in buffer. This dynamic adjustment eliminates unnecessary buffering delays and prevents bandwidth wastage while maintaining efficient frame transfer management.
Data Source
AI summary
There is provided a frame processing device includes a plurality of output ports; a table in which a destination address is stored in association with an output port; a buffer configured to store a learned frame, an un-learned frame, and a copy frame generated by copying the un-learned frame; a transfer unit configured to read a second frame from the buffer in an order in which the second frame is stored and transfer the second frame to a predetermined output port; a storage configured to store the destination address of the learned frame; and a controller configured to discard the second frame to be transferred by the transfer unit, when the second frame is the un-learned frame and the destination address of the second frame is stored in the storage, wherein the second frame transferred to the plurality of output ports is output as the first frame.