Frequency Domain Watermarking for Tabular Data Integrity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data watermarking techniques for tabular data are susceptible to attacks, require a primary key to function effectively, and often damage the data integrity, making them inadequate for robust and non-destructive watermarking.
Innovation Solution
A computer-implemented method for embedding and extracting frequency domain-based watermarks in tabular data, involving the computation of a covariance matrix, projection of row vectors onto orthonormal vectors, determination of a signal space, and embedding or extracting a watermark based on the frequency with the lowest power.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If bit manipulation methods are used for watermarking tabular data, then the watermark can be embedded, but the method is very susceptible to modification and deletion attacks
Solution Approach 1:
The patent transforms the watermarking approach from bit manipulation to frequency domain manipulation. Instead of modifying individual bits in the tabular data, the invention applies Fourier transform to convert the data to frequency domain, then embeds the watermark by modifying frequency coefficients. This parameter transformation makes the watermark robust against attacks because frequency domain modifications are less susceptible to detection and removal compared to bit-level changes.
Solution Approach 2:
The patent replaces the mechanical bit manipulation system with a frequency domain transformation system. By using Fourier transform to convert tabular data into frequency components, the watermarking process substitutes direct bit modification with indirect frequency coefficient adjustment, which provides inherent resistance to modification and deletion attacks.
2Reliability
If statistical methods are used for watermarking, then the watermark can be embedded by altering statistical properties, but the extraction algorithm can fail when data properties are outside assumed ranges
Solution Approach 1:
The patent changes the parameter space from statistical properties (mean, variance) to frequency domain parameters. By transforming data via Fourier transform, the watermark embedding operates on frequency coefficients rather than statistical moments. This approach eliminates the need to assume specific data property ranges, as frequency domain representation is invariant to linear transformations and adapts to any data distribution.
3Reliability
If frequency domain methods are directly applied to tabular data, then watermarking can be performed, but the method severely damages the data impacting its utility
Solution Approach 1:
The patent applies local quality by selectively modifying only the frequency coefficients that correspond to the watermark signal, rather than uniformly transforming all data. The watermark is embedded in specific frequency bands with minimal impact on the overall data structure, preserving data utility while achieving reliable watermarking.
Solution Approach 2:
The patent uses partial action by applying Fourier transform only to the necessary portions of the tabular data and modifying only the specific frequency coefficients required for watermark embedding. This selective approach avoids the excessive transformation that would severely damage data utility, achieving watermarking with minimal data alteration.
Data Source
AI summary
Systems and methods embedding a frequency domain-based watermark in tabular data are provided. In some methods, a covariance matrix is computed for the tabular data. Row vectors of the tabular data are projected onto two orthonormal vectors e1, e2 of the covariance matrix. A signal space for the projected row vectors is determined. A frequency f* with a lowest power from the signal space is located. A watermark is embedded in the tabular data based on the frequency f*.