Time-Controlled Message Distribution in Distributed Sensor Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In distributed real-time computer systems, the transmission of real-time data from hundreds of spatially distributed sensors to a control center is inefficient due to the high latency and data volume, especially in fast technical processes like Smart Grids, where standard protocols result in excessive data overhead and prolonged transmission times.
Innovation Solution
The system employs a topology structure resembling an in-tree, where node computers at the leaves send sensor data in time-controlled real-time messages to a control center, and distribution units at inner nodes copy payloads from multiple senders into a single new message, optimizing transmission by creating individual schedules for each distribution unit to ensure timely arrival of data at the control center.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If standard protocols like TT Ethernet are used for transmitting sensor data, then data transmission reliability is ensured, but transmission time increases significantly due to excessive data overhead
Solution Approach 1:
The patent segments the Ethernet message into two parts: a standard header transmitted using reliable TT Ethernet protocol, and a payload containing only essential sensor data. This segmentation allows the critical time-sensitive data to be transmitted with minimal overhead while maintaining protocol reliability for the header portion.
Solution Approach 2:
The patent extracts only the necessary sensor data from the complete sensor information and transmits it in the payload, leaving out non-critical data. This extraction reduces the payload size significantly, thereby reducing total transmission time while maintaining data transmission reliability through the standardized header.
2Reliability
If sensor data from multiple sensors is transmitted separately in individual Ethernet messages, then data integrity is maintained, but total data volume and transmission time increase
Solution Approach 1:
The patent merges payloads from multiple Ethernet messages containing sensor data from different sensors into a single combined payload. This merging reduces the total number of messages transmitted and decreases overall data volume, while the standardized header structure ensures data integrity is maintained through consistent formatting and error checking.
3Adaptability or versatility
If minimum Ethernet message length of 64 bytes is used for each sensor data transmission, then protocol compliance is ensured, but transmission efficiency decreases due to excessive overhead
Solution Approach 1:
The patent segments the message structure into a standardized header (ensuring protocol compliance) and a compact payload (optimizing efficiency). The header maintains the minimum 64-byte Ethernet message requirement for protocol compliance, while the payload contains only essential sensor data, improving transmission efficiency by eliminating redundant data.
Solution Approach 2:
The patent changes the parameter of payload size to be minimal and optimized for the specific sensor data being transmitted. By adjusting the payload content to include only necessary information rather than fixed-size data fields, the system maintains protocol compliance while significantly improving transmission efficiency through reduced data volume.
4Reliability
If data from hundreds of sensors is transmitted simultaneously, then comprehensive process monitoring is achieved, but network latency increases
Solution Approach 1:
The patent merges sensor data from hundreds of sensors into consolidated payloads at distribution units, reducing the total number of messages that need to traverse the network. This merging maintains comprehensive process monitoring by preserving all sensor data while reducing network latency through fewer transmission events and optimized data volume.
Solution Approach 2:
The patent introduces distribution units as intermediaries between sensors and the control center. These intermediaries aggregate and consolidate sensor data from multiple sources before transmission, achieving comprehensive process monitoring while reducing network latency by minimizing the number of individual message transmissions across the network.
Data Source
Figure 1~2
Figure 3
AI summary
The invention relates to a method, in particular a time-controlled error-tolerant method, for periodically transporting real-time data in a computer system, in particular in a distributed computer system, said computer system comprising node computers (111 - 116), in particular a plurality of node computers (111 - 116), and distributor units (131, 132, 133, 151), in particular a plurality of distributor units (131, 132, 133, 151). The node computers and the distributor units have access to a global time, and real-time data is transported by means of messages, preferably by means of time-controlled real-time messages. The topology of the computer system corresponds to an intree, and node computers (111 - 116) are arranged on the leaves of the intree. One or more sensors (101 - 106) is assigned to each node computer (111 - 116), and the node computers (111 - 116) arranged on the leaves of the intree transmit sensor data in the payload of messages in the direction of a control center (100) located at the root of the intree at points in time, preferably at synchronized points in time. The payload of one or more incoming messages in a distributor unit is transported out of the distributor unit with an outgoing message, and an individual time plan is generated a priori for each distributor unit, wherein the time plan contains a periodically repeating starting time (391) for transmitting a message (390) going out of the respective distributor unit, said starting time being calculated a priori from - the a priori known time of arrival of a controlling payload to be transported, which is specified a priori for example, of one of the incoming messages, in particular a time-critical payload to be transported of one of the incoming messages, - minus the a priori known lead-time interval ([391, 393]) of the outgoing message (390), - plus at least one time interval ([393, 393)] required to copy a data element of the controlling payload, in particular the time-critical payload of the incoming message (320), into a data element of the payload of the outgoing message (390); and the payloads of the incoming messages are copied into the payload of the outgoing message by carrying out the a priori generated time plan.