Packet Analysis Device Flow Control Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Passive network bandwidth measurement is challenging due to difficulties in identifying continuous packets affected by flow control, leading to inaccurate bandwidth estimation.
Innovation Solution
A packet analysis method that acquires acknowledge packets and calculates bandwidth values based on the timing of packet acquisitions, identifying and removing packets affected by flow control to determine accurate network bandwidth.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If passive diagnosis is used to measure network bandwidth, then no diagnosis packets need to be prepared and network congestion is avoided, but accurate bandwidth measurement becomes difficult due to flow control delays
Solution Approach 1:
The patent applies preliminary action by pre-calculating expected packet arrival times based on transmission intervals before actual measurement. The measurement device compares actual packet arrival timing against these pre-established expectations to identify packets delayed by flow control, enabling accurate bandwidth measurement without active diagnosis packets.
Solution Approach 2:
The patent substitutes mechanical timing analysis with information processing by using packet sequence numbers and transmission interval metadata. Instead of relying solely on physical packet arrival timing, the system uses informational cues (sequence numbers, expected intervals) to identify flow-controlled packets, replacing direct temporal measurement with computational analysis.
2Measurement precision
If packets transmitted immediately after ACK reception are identified for bandwidth measurement, then flow control impact is captured, but time difference between terminal transmission and observation point reception makes identification difficult
Solution Approach 1:
The patent introduces packet sequence numbers and expected transmission interval metadata as intermediaries between the terminal and observation point. These informational mediators allow the observation point to reconstruct transmission timing expectations without directly observing terminal transmission moments, bridging the time difference gap through information rather than direct temporal correlation.
Solution Approach 2:
The patent transitions from one-dimensional temporal analysis to multi-dimensional analysis by incorporating packet sequence numbers and expected transmission intervals as additional dimensions. Instead of relying solely on arrival time, the system uses sequence number progression and interval metadata to identify flow-controlled packets, adding informational dimensions to compensate for temporal displacement.
Data Source
AI summary
A packet analysis method includes acquiring a first acknowledge packet and a second acknowledge packet transmitted from the first device; acquiring a plurality of packets transmitted from the second device during a period from reception of the first acknowledge packet to reception of the second acknowledge packet; identifying a number of packets corresponds to data transmitted from the second device, by calculating a difference between a first identification number corresponding to the first acknowledge packet and a second identification number corresponding to the second acknowledge packet; calculating a plurality of bandwidth values, a number of the plurality of bandwidth values is identical to the number of packets, based on acquisition timings of the plurality of packets and an interval between the acquisition timings of two adjacent packets; and determining a bandwidth value that is to be removed from the plurality of bandwidth values by comparing the plurality of bandwidth values.