Network Switch CAM Traffic Monitoring Padding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Content addressable memory (CAM) systems, particularly ternary content addressable memory (TCAM), face challenges in handling data transmissions of sizes that do not match the exact size or integer multiples of the CAM rules, leading to incorrect denial of data transmissions due to size mismatches and increased latency when evaluating larger data packets.
Innovation Solution
The method involves padding data transmissions to match the CAM rule size by adding predetermined data and assigning identification numbers to data packets, allowing the CAM to compare and process data efficiently, while the CPU evaluates only the remaining data, thereby preventing denial due to size issues and reducing latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If data transmissions are directly compared with CAM rules without size matching, then processing speed is maintained, but data transmissions with sizes not matching CAM rule sizes are incorrectly denied
Solution Approach 1:
The patent applies preliminary action by padding data transmissions with predetermined data before CAM comparison to ensure size matching. This preprocessing step ensures that data transmissions of any size are extended to match CAM rule sizes (e.g., 64 bytes), preventing incorrect denial while maintaining fast CAM-based processing. The padding occurs before the comparison operation, allowing the CAM to process all data uniformly at high speed without rejecting valid transmissions due to size mismatches.
2Measurement precision
If the CPU evaluates all data in data transmissions, then comprehensive evaluation is achieved, but latency increases for larger data packets
Solution Approach 1:
The patent applies segmentation by dividing data transmission evaluation into two parts: CAM handles the fixed-size portion (e.g., first 64 bytes containing header information) while the CPU evaluates only the remaining variable-size payload. This segmentation allows the CAM to quickly filter and classify the structured portion of data transmissions, then pass only the necessary remaining data to the CPU for detailed evaluation, significantly reducing CPU latency for large packets while maintaining comprehensive evaluation of all relevant data.
3Quantity of substance
If CAM rules are made larger to accommodate more data, then more data can be processed, but the complexity and resource requirements increase
Solution Approach 1:
The patent applies partial action by having the CAM process only a fixed, optimized portion of data transmissions (e.g., 64 bytes containing essential header information) rather than requiring CAM rules to accommodate entire variable-size packets. This allows the use of smaller, more manageable CAM rules that can be quickly searched, while the CPU handles evaluation of the remaining data. This approach processes the critical portion of all packets through the fast CAM path without requiring excessive CAM rule sizes.
Data Source
Figure 1
Figure 2
AI summary
The present invention relates to a method for monitoring traffic between network members in a network, especially in an automotive network, wherein in the network different network members are connected with a network switch comprising a processor unit and a content addressable memory CAM, in which rules concerning data transmissions between different network members are stored; wherein upon receiving (205) data to be transmitted between two network members the data to be transmitted are compared (208) with the rules stored in the CAM and depending on a result of this comparison it is selected whether to allow (209) the data transmission or to deny (209) the data transmission or to transfer (210) the data to be transmitted to the processor unit; wherein if the data to be transmitted have a size smaller than a first threshold depending on a size of the rules stored in the CAM, predetermined data are added (207) to the data to be transmitted such that the data to be transmitted have a size equal to or greater than the first threshold and afterwards the data to be transmitted are compared (208) with the rules stored in the CAM; and/or wherein if the data to be transmitted have a size larger than a second threshold depending on the size of the rules stored in the CAM and if it is selected (210) to transfer the data to be transmitted to the processor unit, a predetermined identification number is assigned (214) to the data to be transmitted.