Locating FPDU Headers in TCP Segments Without Markers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current protocols face challenges in locating protocol data unit (PDU) headers within TCP segments that arrive out of order without relying on cumbersome markers, which increases hardware complexity and requires additional computing power for marker insertion and extraction.
Innovation Solution
The method involves performing cyclic redundancy checks (CRC) on TCP segments to identify potential PDU headers, walking payload testing key fields in the DDP header to determine the offset probability, and verifying the actual PDU through further CRCs, allowing for the processing of PDUs without markers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If markers are used to locate FPDU headers in out-of-order TCP segments, then the receiving node can easily identify PDU headers, but hardware complexity increases and additional computing power is required for marker insertion and extraction
Solution Approach 1:
The patent extracts the header location information directly from the existing TCP segment data by analyzing payload testing key fields and performing CRC checks, rather than adding separate marker elements. This removes the need for additional marker insertion/extraction hardware while maintaining the ability to locate headers efficiently
Solution Approach 2:
The patent uses CRC checks and payload testing key field analysis as intermediary verification mechanisms to identify header locations without requiring physical markers. These intermediaries provide a computational pathway to locate headers that avoids the hardware complexity of marker-based systems
2Difficulty of detecting and measuring
If markers are used to locate FPDU headers in out-of-order TCP segments, then the receiving node can easily identify PDU headers, but additional computing power is required for marker insertion and extraction
Solution Approach 1:
The patent extracts header location information from existing data fields (payload testing key fields, CRC values) rather than performing computations on added marker elements. This reduces computational overhead by utilizing already-present data structures to determine header positions
Solution Approach 2:
The TCP segment data itself contains the information needed to locate headers through its existing structure (CRC fields, payload key fields). The system uses self-contained data within the segment to identify headers, eliminating the need for external marker processing and reducing computational requirements
3Device complexity
If DDP segments are sent without markers, then hardware complexity is reduced, but the receiving node must perform multiple CRC checks and walk payload fields to locate FPDU headers
Solution Approach 1:
The patent performs preliminary analysis of payload testing key fields and CRC values at the receiving node to quickly identify potential header locations before full processing. This preliminary action uses readily available data structures to narrow down search space, reducing the time penalty of not using markers
Data Source
AI summary
The present system and methods perform a series of cyclic redundancy checks (CRCs) on data within a transmission control protocol (TCP) segment. If any of the CRCs are successful, it is determined that the FPDU has been located, and the FPDU is processed. When one of the CRCs fails, additional steps are taken to determine if a potential FPDU is in fact an FPDU. In one such step a plurality of payload testing key fields in a direct data placement (DDP) header of an offset in the TCP segment are walked to determine a probability of the offset being the start of an FPDU.