Dual Priority Flow Control for Packet Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing flow control methods in communication technologies result in packet loss and network congestion when packets are switched between priority queues, due to inaccurate flow control information and queue scheduling.
Innovation Solution
A method where packets carry two priorities, one for flow control and another for queue scheduling, allowing for accurate flow control information to be sent upstream, preventing packet loss and congestion by ensuring correct queue placement and flow control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a packet is switched between priority queues to resolve network deadlock or congestion, then network flow control is improved, but packet loss occurs due to inaccurate flow control information
Solution Approach 1:
The patent segments the priority field into two distinct parts: a first priority value used for flow control decisions and a second priority value used for queue scheduling. This segmentation allows the system to independently optimize flow control accuracy while maintaining proper queue management, thereby preventing packet loss during priority queue switching operations.
Solution Approach 2:
The patent introduces an intermediary mechanism where the first priority value acts as a mediator between the original packet priority and the flow control system. This intermediary priority value enables accurate flow control information transmission to upstream devices without affecting the actual queue scheduling based on the second priority value.
2Productivity
If credit-based flow control is used with priority queue switching, then buffer management is improved, but packet accumulation occurs due to incorrect credit value transmission
Solution Approach 1:
The patent segments the priority information into two separate values: the first priority value is used for credit-based flow control and credit value transmission, while the second priority value is used for queue scheduling. This ensures that credit values are transmitted accurately according to the flow control priority rather than the queue scheduling priority, preventing packet accumulation caused by incorrect credit management.
Solution Approach 2:
The patent implements feedback mechanisms where upstream devices receive accurate flow control information based on the first priority value, allowing them to adjust their transmission rates appropriately. This feedback loop prevents packet accumulation by ensuring that credit values and flow control signals accurately reflect the actual buffer status and flow control requirements.
3Device complexity
If single priority field is used for both flow control and queue scheduling, then device complexity is reduced, but network congestion occurs due to inaccurate flow control information
Solution Approach 1:
The patent divides the priority field into two separate priority values: the first priority value dedicated to flow control operations and the second priority value dedicated to queue scheduling. This segmentation resolves the conflict between device simplicity and network efficiency by allowing each priority value to independently optimize its specific function, thereby preventing network congestion while maintaining manageable device complexity.
Solution Approach 2:
The patent applies local quality by assigning different functions to different parts of the priority information structure. The first priority value is optimized for flow control accuracy, while the second priority value is optimized for queue scheduling. This localized optimization ensures that each aspect of packet handling operates with the appropriate priority information, improving overall network efficiency without excessive complexity.
Data Source
AI summary
Embodiments of this application relate to the field of communications technologies, and disclose a flow control method and apparatus, to resolve a prior-art problem such as packet loss, packet accumulation, or network congestion that occurs after a packet is switched between priority queues. A specific solution is as follows: A first device receives a first packet sent by a second device, where the first packet carries a first field and a second field, the first field carries a first priority, and the second field carries a second priority; the first device performs flow control based on the first priority in the first packet; and the first device performs queue scheduling on the first packet based on the second priority in the first packet.