Coded Data Processing with DNB Signatures for Lower Runtime Overhead

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

ANBD coding for data processing is associated with high runtime performance overhead and requires additional data, necessitating complex hardware to handle encoded data effectively.

Innovation Solution

The proposed method uses DNB coding, which simplifies the coding process by eliminating the input constant A, reducing computational effort and hardware requirements by up to 40% compared to ANBD coding, while maintaining performance through cycle-specific and variable-specific signatures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If ANBD coding is used for data processing, then data can be protected from errors and undetected modifications, but runtime performance overhead increases and additional supplementary data is required

Engineering Contradiction:
Improvedata protectionVSAvoidruntime performance
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent extracts and removes the input constant A from the ANBD coding formula, creating a simplified DNB coding method that uses only the cycle-specific signature D and variable-specific signature Bx. This extraction eliminates unnecessary computational overhead while preserving the essential error detection functionality through the remaining signature components.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the parameters of the coding formula by eliminating the input constant A and modifying the coding equation from xc = (A + Bx · x + D) modulo 2^n to xc = (Bx · x + D) modulo 2^n. This parameter change reduces computational complexity from multiple arithmetic operations to simpler operations, improving runtime performance while maintaining data protection capabilities.

Inventive Principle:
Principle #35Parameter changes

2Reliability

If ANBD coding is used for data processing, then data can be protected from errors, but additional supplementary data is required for transmission and storage

Engineering Contradiction:
Improveerror detectionVSAvoiddata overhead
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent extracts and removes the input constant A from the coding process, eliminating the need to transmit or store this additional supplementary data. The DNB coding method achieves error detection functionality using only the cycle-specific signature D and variable-specific signature Bx, reducing the quantity of data required for transmission and storage.

Inventive Principle:
Principle #2Taking out (Extraction)

3Reliability

If ANBD coding is used for data processing, then coded data can be processed, but sophisticated hardware with powerful CPUs is needed

Engineering Contradiction:
Improvecoded data processingVSAvoidhardware requirements
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent changes the computational parameters of the coding method by eliminating the input constant A and simplifying the arithmetic operations required. This parameter change reduces the computational burden from complex multi-step arithmetic operations to simpler operations that can be performed with less sophisticated hardware, reducing device complexity while maintaining coded data processing capabilities.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent extracts and removes the computationally intensive input constant A from the processing workflow, eliminating the need for powerful CPUs to handle complex calculations. The simplified DNB coding can be executed on less sophisticated hardware while still providing reliable coded data processing and error detection.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP3614268B1Method and device for processing of data using coded operations
Publication Date: 2021.07.07 SIEMENS AG
  • EP3614268B1 patent drawingFigure 1~2
  • EP3614268B1 patent drawingFigure 3
  • EP3614268B1 patent drawing

AI summary

A method for processing data using coded operations in a number of cycles is proposed. An uncoded variable x is coded to a coded variable Xc using a cycle-specific signature D and a variable-specific signature Bx, according to the equation xc = D ⋅ x + Bx. Furthermore, a computer program, a device, and a functionally safe, programmable controller are proposed.