Packet Congestion Processing via Fragment Parameter Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current packet congestion processing methods, such as Weighted Random Early Detection (WRED), face high energy consumption and long waiting times due to the need for RAM storage of packet fragments, which becomes unsustainable with a large number of network interfaces.
Innovation Solution
A method and device that process packet congestion by receiving packet fragments and fragment parameters, calculating the average queue size and size thresholds, and applying a WRED policy to determine whether to discard or pass packets, thereby reducing RAM usage and accelerating processing speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If packet fragments are stored in RAM to acquire packet size for congestion processing, then packet congestion can be processed using WRED algorithm, but RAM resource consumption increases proportionally with the number of network interfaces
Solution Approach 1:
The patent extracts only the necessary packet size information from incoming packet fragments without storing the entire packet fragments in RAM. By calculating packet size on-the-fly from fragment parameters and extracting only essential metadata, the system maintains congestion processing capability while dramatically reducing RAM consumption proportional to the number of interfaces.
Solution Approach 2:
The patent performs preliminary calculations of packet size and queue metrics as packet fragments arrive, rather than waiting to store complete packets. By computing congestion parameters incrementally during fragment reception, the system avoids the need to hold entire packets in memory while maintaining accurate congestion state information.
2Measurement precision
If packet fragments are stored in RAM until entire packet is received, then accurate packet size can be obtained for congestion processing, but processing waiting time increases
Solution Approach 1:
The patent calculates packet size metrics and queue parameters as soon as fragment parameters are received, performing preliminary congestion assessment without waiting for the complete packet. This allows the system to make congestion processing decisions based on accurate size estimates while minimizing waiting time by not requiring full packet reception before processing.
Solution Approach 2:
The patent skips the traditional step of storing and waiting for complete packet reception by rushing through the congestion processing using available fragment information. The system rapidly computes queue sizes and makes discarding decisions based on partial information, eliminating the time-consuming waiting period while maintaining measurement accuracy through proper parameter tracking.
3Productivity
If WRED algorithm is used with average queue size calculation, then network throughput can be maximized, but energy consumption increases due to continuous queue monitoring and packet fragment storage
Solution Approach 1:
The patent extracts only the essential queue size metrics needed for WRED decision-making without storing complete packet fragments. By calculating and maintaining only the necessary average queue size parameters through incremental updates as fragments arrive, the system preserves network throughput optimization while eliminating the energy-intensive operation of storing and managing complete packet fragment data in RAM.
Data Source
AI summary
Provided are a method and device for processing packet congestion. The method includes: receiving a packet fragment of a packet and fragment parameters and acquiring congestion parameters required for processing packet congestion according to the fragment parameters; when the packet fragment is an SOP fragment in the packet, acquiring an average queue size and a size threshold of the packet according to the fragment parameters and the congestion parameters and processing the packet. By acquiring the average queue size and the size threshold of the packet according to the fragment parameters and congestion parameters received and processing the packet according to the relationship between the average queue size and the size thresholds when a packet fragment received is an SOP fragment, the solution effectively eliminate the packet congestion in a communication network, accelerate the operation speed of processing the packet congestion and greatly save an internal cache space.