Unidirectional Data Transfer With Segmented Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In safety-critical industrial networks, data transfer to open networks is often unidirectional and limited by devices with constrained computing and main-memory capacity, where existing forward error correction methods increase transfer volumes and require significant memory resources for decoding.
Innovation Solution
A method and system for error-correcting data transmission that divides data records into subrecords, adds correction characters, and transmits them via a unidirectional communication unit, allowing iterative decoding and storage in devices with limited resources by using a main memory for initial subrecords and a background memory for others, with a management unit to optimize character and correction character allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If forward error correction methods are used to protect data transfer, then data transmission reliability is improved, but transfer volume increases and memory requirements increase
Solution Approach 1:
The data record is divided into multiple data subrecords, each processed and transmitted separately. This segmentation allows the receiving device to decode and store subrecords individually in main memory rather than holding entire coded data records, reducing peak memory requirements while maintaining error correction capabilities through distributed processing
2Reliability
If forward error correction methods are used to protect data transfer, then data transmission reliability is improved, but device memory capacity requirements increase
Solution Approach 1:
By segmenting the data into subrecords and processing them individually, the system reduces the amount of data that must be held in main memory simultaneously. The receiving device only needs to store one decoded subrecord in main memory at a time, while using background memory for buffer storage of additional subrecords
Solution Approach 2:
The invention introduces a temporal dimension to memory usage by iteratively decoding subrecords and moving them from background memory to main memory as needed. This allows the system to manage limited main memory resources efficiently by processing data in sequences rather than requiring all data to be available simultaneously
3Device complexity
If iterative decoding with background memory is used, then device complexity is reduced, but processing time increases
Solution Approach 1:
The receiving device pre-allocates background memory for buffer storage of coded data subrecords before decoding begins. This preliminary preparation allows the iterative decoding process to proceed efficiently without requiring complex dynamic memory management decisions during the actual decoding operations
Solution Approach 2:
The system dynamically manages memory allocation by iteratively moving decoded subrecords from background memory to main memory and loading new subrecords into background memory as needed. This dynamic approach allows flexible adaptation to varying data sizes and decoding requirements while maintaining simple overall device architecture
Data Source
AI summary
A method and a system for fault-correcting transfer of a dataset from a first network into a second network via a unidirectional communication unit, the receiving apparatus having a limited computing and main memory capacity is provided. The dataset is divided into partial datasets and are each coded by adding at least one correction mark. The marks and correction marks are decoded repeatedly in the second network, wherein marks and correction marks associated with a first encoded partial dataset are stored in a main memory, marks and correction marks associated with other encoded partial datasets are buffered in a background memory, and the first coded partial dataset is decoded, and then the previously received marks and/or correction marks of one of the other coded partial datasets are shifted from the background memory into the main memory. The dataset is reproduced from the decoded partial datasets in the second network.