Network Device CRC Error Source Identification via Packet Duplication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In computer networks using cut-through forwarding, corrupted network packets can propagate across multiple devices before errors are detected, making it difficult to identify the source of cyclic redundancy check (CRC) errors due to the nature of packet forwarding where CRC errors are only verified at the end of the packet, and resources are wasted processing and re-transmitting erroneous packets.
Innovation Solution
Duplicating a portion of incoming network packets and processing them through a secondary pipeline that performs CRC checks, allowing for the identification of the source device even when cut-through forwarding is used, by maintaining packet headers and performing CRC checks similar to store-and-forward processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If cut-through forwarding is used to reduce latency, then network packet transmission speed is improved, but the ability to detect and identify CRC errors deteriorates
Solution Approach 1:
The network device is divided into two separate forwarding paths: a first forwarding path that performs cut-through forwarding for low-latency transmission, and a second forwarding path that performs store-and-forward processing for reliable CRC error detection. This segmentation allows both speed and reliability requirements to be satisfied simultaneously by routing different packet copies through different paths.
Solution Approach 2:
The patent creates a duplicate copy of the network packet and processes it through a separate store-and-forward path. The original packet travels through the fast cut-through path while the copied packet travels through the reliable store-and-forward path for CRC verification, enabling error detection without compromising the speed of the primary transmission path.
2Loss of time
If cut-through forwarding is used, then latency is reduced, but the ability to identify the source of corrupted packets deteriorates
Solution Approach 1:
The patent segments the packet processing into two independent paths: the first path provides fast cut-through forwarding with minimal latency, while the second path performs complete packet reception and source identification. This allows the system to maintain low latency for transmission while separately performing accurate source tracking.
Solution Approach 2:
A duplicate packet is created and processed through the store-and-forward path where the complete packet including source address information is received and recorded. This copied packet serves as a reference for identifying the source of corrupted packets without delaying the original packet's fast transmission.
3Reliability
If store-and-forward processing is used for CRC checks, then error detection reliability is improved, but network packet transmission speed deteriorates
Solution Approach 1:
The patent divides packet processing into two specialized paths: a cut-through path optimized for speed that forwards packets immediately upon receiving the header, and a store-and-forward path optimized for reliability that completes full packet reception and performs thorough CRC verification. Each path is segmented to perform its specific function efficiently.
Solution Approach 2:
The network device implements multi-functionality by handling both cut-through forwarding and store-and-forward processing within the same device. The system can selectively route packets through different paths based on requirements, making the device universally capable of both high-speed transmission and high-reliability error detection.
Data Source
AI summary
Technologies are provided for identifying the source of cyclic redundancy check (CRC) errors within a computer network that uses cut-through forwarding of network packets. For example, inbound network packets can be received by a network device that is configured to perform cut-through forwarding. Once the network packets are received, they are processed using cut-through forwarding. In addition to the cut-through forwarding, at least some of the received network packets are duplicated to create duplicate network packets. The duplicate network packets are processed using a different procedure. For example, the duplicate network packets are processed (e.g., as part of a secondary processing pipeline) by performing CRC checks on the duplicate network packets to detect CRC errors. Based on the CRC checks, source devices that originated network packets with CRC errors can be identified.