Network Communication Apparatus with Parallel Serial Core Groups
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network communication apparatuses face high memory and computing power requirements due to the need for packet processing in tunnel networks, particularly for MTU mechanisms, leading to a heavy workload.
Innovation Solution
A network communication apparatus with a 'step-based multi-processor' architecture and 'shared memory' configuration, featuring a dispatch device and core groups with parallel and serial core units, which offloads packet processing efforts and balances workload through meta data dispatch and processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If packet processing is performed for the MTU mechanism in tunnel networks, then packets can be transmitted between private networks across public networks, but the network communication apparatus requires more memory space and computing power resulting in heavy workload
Solution Approach 1:
The network communication apparatus is divided into multiple core groups (first core group with parallel core units, second core group with serial core units) that process packets in stages. The dispatch device distributes packets to different core groups based on processing requirements, segmenting the overall packet processing workload across multiple specialized processing units rather than concentrating all processing in a single apparatus.
Solution Approach 2:
A dispatch device is introduced as an intermediary component that receives packets from the external interface and intelligently distributes them to appropriate core groups. The dispatch device contains a determination module that decides whether packets require MTU mechanism processing, and a distribution module that routes packets to either parallel or serial core groups based on the processing needs, thereby mediating between packet input and processing resources.
2Adaptability or versatility
If packet processing is performed for the MTU mechanism, then tunnel network communication is enabled, but more memory space and computing power are required
Solution Approach 1:
Different core groups are assigned different processing qualities based on packet requirements. The first core group with parallel core units handles packets requiring MTU mechanism processing (deencapsulation, fragmentation), while the second core group with serial core units handles packets that only require encapsulation. This local quality differentiation ensures that computing power and memory space are consumed only where necessary for tunnel network communication.
3Device complexity
If a single core unit processes all packets, then processing is simplified, but processing speed and throughput are limited
Solution Approach 1:
The system dynamically adjusts processing architecture by using parallel core units for packets requiring intensive MTU processing and serial core units for simpler encapsulation tasks. The dispatch device dynamically routes packets based on real-time processing requirements, enabling the system to adapt its processing speed and resource utilization to match the actual workload characteristics.
Solution Approach 2:
The patent transitions from a single-dimension serial processing model to a multi-dimensional processing architecture by introducing both parallel and serial core groups operating simultaneously. This dimensional change allows packets to be processed in parallel when needed (increasing throughput) while maintaining serial processing simplicity for other cases, thereby improving overall packet processing speed without excessive complexity.
Data Source
AI summary
A network communication apparatus, includes a dispatch device, a first core group with several parallel core units and a second core group with at least one serial core unit. The dispatch device receives several packets contained in several first packet flows, and configured to dispatch several meta data to the parallel core units through several first data flows, and the meta data contain tunnel parameters of the packets. Furthermore, the at least one serial core unit receives the meta data from the parallel core units through several second data flows.