PROFINET Diagnostic Packet Flooding for Fault Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
PROFINET RT networks face difficulties in diagnosing transmission faults, particularly when non-real-time communication bursts cause real-time frame losses, leading to system downtime and high diagnostic costs due to sluggish standard diagnostic tools like SNMP, which fail to effectively identify the source of issues within the network.
Innovation Solution
A method involving targeted flooding of diagnostic data packets of the lowest priority across the network to achieve 100% utilization, allowing immediate detection of packet losses and identification of the source of interference through unique counters and analysis of missing packets, thereby quickly pinpointing the cause of disruptions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard diagnostic tools like SNMP are used to monitor network transmission faults, then the network can be monitored with standard hardware and software, but the diagnostic response is too sluggish to effectively identify the source of issues in real-time
Solution Approach 1:
The patent introduces diagnostic data packets that continuously circulate through the network beforehand, so that when a fault occurs, the system can immediately detect which packets are missing or delayed. This preliminary placement of diagnostic probes throughout the network eliminates the need for reactive troubleshooting, providing both real-time detection (solving the time loss problem) and precise fault localization (solving the measurement precision problem).
Solution Approach 2:
The patent uses special diagnostic data packets as intermediaries to carry timing and identification information through the network. These packets act as mediators between the network components and the diagnostic system, enabling precise measurement of packet transmission times and identification of fault locations without requiring modification of the actual production data traffic.
2Adaptability or versatility
If NRT communication bursts are allowed in the network to maintain flexibility and compatibility with standard Ethernet, then network adaptability is improved, but RT frames are discarded due to queue overflow in network elements
Solution Approach 1:
The patent implements a feedback mechanism where diagnostic data packets continuously monitor network conditions and provide information about packet loss and transmission delays. This feedback enables the system to detect when NRT bursts are causing RT frame discards, allowing for dynamic adjustment of network parameters or identification of problematic NRT sources while maintaining overall network adaptability.
3Quantity of substance
If network logs are recorded over long periods to capture intermittent faults, then more diagnostic data is collected, but the burst patterns are smeared across multiple records and become indistinguishable
Solution Approach 1:
The patent uses periodically circulating diagnostic data packets that traverse the network at regular intervals. Each packet carries timestamp and identification information that allows precise correlation of packet loss events with specific time windows. This periodic action maintains high measurement precision even over long monitoring periods, avoiding the smearing effect of continuous logging by using discrete, time-stamped diagnostic probes.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
When using PROFINET IRT in a system, the transmission of process data is determined ahead of time in an extremely precise manner for the entire network according to a precisely determined sequence using an engineering tool. The maximally allowed deviation from the beginning of a bus cycle is 1 µs, thereby resulting in no queues during communication. The preplanning in the IRT component and the monitoring process carried out by the IRT hardware excludes the possibility of the NRT protocol interfering with the IRT communication component. The PROFINET standard requires that all participants do not transmit more than 3 Kbyte/ms. Such a requirement does not exist in IEEE 802.3, however. More often, there are multiple extensions which oppose said requirement, thereby leading to RT frames now being discarded due to overflowing queues in the network elements because the network component is blocked by NRT communication jumbo frames that are being transmitted at the moment for example. Thus, the RT communication can no longer be maintained such that as a result, the subordinate automation task can no longer be carried out in an error-free manner. In order to diagnose the problem, the network is flooded with packets of the lowest priority so that the network is being utilized at virtually 100% capacity. Said packets "know" their path and allow a determination to be made as to which and how many data packets have been lost. By further analyzing the missing diagnosis data packets, a detection of the network components in which the packet loss has occurred can be carried out.