Packet Acceleration via Custom Connection Tracking Table
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing 464XLAT packet translation technology experiences high CPU occupation and low transmission rates due to complex processing logic in the standard TCP/IP protocol stack, leading to poor user experience.
Innovation Solution
A packet transmission method that identifies packets requiring acceleration and queries a customized connection tracking table to acquire connection tracking information, bypassing the complex standard TCP/IP protocol stack processing, thereby reducing CPU occupation and increasing transmission rates without the need for hardware upgrades.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the standard TCP/IP protocol stack processing is used for packet translation, then the translation function is complete, but the CPU occupation rate is high and transmission rate is low
Solution Approach 1:
The patent segments packet processing into two paths: standard TCP/IP protocol stack processing for normal packets, and a simplified customized connection tracking table processing path for packets requiring acceleration. This segmentation allows critical translation functions to be completed while avoiding unnecessary complex processing for time-sensitive packets, thereby improving transmission rate without sacrificing translation completeness.
Solution Approach 2:
The patent extracts the connection tracking function from the standard TCP/IP protocol stack and implements it independently through a customized connection tracking table. By taking out this specific function, the system can process packets requiring acceleration through a simplified path that bypasses the complex standard protocol stack, reducing CPU occupation and improving transmission rate.
2Ease of operation
If the standard TCP/IP protocol stack processing is used, then all packets are processed uniformly, but the CPU occupation rate increases
Solution Approach 1:
The patent introduces dynamic processing paths based on packet characteristics. Packets are classified into different processing modes: those requiring acceleration use the simplified customized connection tracking table path, while other packets use the standard TCP/IP protocol stack path. This dynamic approach maintains processing uniformity for packets that need it while reducing CPU occupation for packets that can use the accelerated path.
Solution Approach 2:
The patent applies different processing qualities to different packets based on their requirements. For packets requiring acceleration, a simplified processing quality is applied through the customized connection tracking table, while other packets receive full standard protocol stack processing. This local differentiation reduces overall CPU occupation while maintaining necessary processing quality where needed.
Data Source
AI summary
Provided are a packet transmission method and device and a computer readable storage medium. The packet transmission method includes: receiving, by the routing device, a packet, determining whether the packet is a packet that needs to be accelerated, in response to determining that the packet is the packet that needs to be accelerated, querying a customized connection tracking table to acquire connection tracking information corresponding to the packet, transmitting the packet to a driver layer for processing, where the customized connection tracking table includes connection tracking information of a packet corresponding to a packet type which is predetermined to be accelerated.