SSO Noise Prediction Using Superposition and Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for predicting simultaneous switching output (SSO) noise in integrated circuits, such as FPGAs, face challenges due to complex 3-dimensional structures and computationally intensive algorithms, leading to limited accuracy and high computational resource requirements.
Innovation Solution
A method using superposition and pre-recorded noise waveforms to predict SSO noise by selecting victim and aggressor signal paths, recording electrical disturbances, and aggregating them to predict output noise, which reduces computational complexity and improves accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional 3-dimensional modeling tools are used to predict SSO noise, then measurement precision is improved, but device complexity and computational resource requirements increase significantly
Solution Approach 1:
The patent segments the complex 3D electromagnetic coupling problem into multiple 2D cross-sectional planes. Each plane is analyzed independently to extract coupling coefficients, which are then synthesized to predict SSO noise. This segmentation reduces computational complexity while maintaining prediction accuracy by avoiding the need for full 3D electromagnetic simulations.
Solution Approach 2:
The patent performs preliminary extraction of electromagnetic coupling coefficients from 2D cross-sections before conducting SSO noise prediction. These pre-computed coefficients are stored and reused for multiple noise scenarios, eliminating the need to repeatedly solve complex electromagnetic field equations and significantly reducing computational resources required for actual noise prediction.
2Measurement precision
If conventional 3-dimensional modeling tools are used to predict SSO noise, then measurement precision is improved, but productivity decreases due to high computational resource requirements
Solution Approach 1:
The patent segments the complex 3D electromagnetic coupling problem into multiple 2D cross-sectional planes. Each plane is analyzed independently to extract coupling coefficients, which are then synthesized to predict SSO noise. This segmentation reduces computational complexity while maintaining prediction accuracy by avoiding the need for full 3D electromagnetic simulations.
Solution Approach 2:
The patent performs preliminary extraction of electromagnetic coupling coefficients from 2D cross-sections before conducting SSO noise prediction. These pre-computed coefficients are stored and reused for multiple noise scenarios, eliminating the need to repeatedly solve complex electromagnetic field equations and significantly reducing computational resources required for actual noise prediction.
3Device complexity
If approximations are used to reduce computational complexity, then device complexity is reduced, but measurement precision deteriorates due to inaccurate representation of electrical principles
Solution Approach 1:
The patent introduces electromagnetic coupling coefficients as intermediary parameters that bridge the gap between simple 2D cross-sectional analysis and complex 3D electromagnetic behavior. These coefficients capture the essential electromagnetic coupling characteristics without requiring full 3D simulations, enabling accurate SSO noise prediction with reduced computational complexity.
Solution Approach 2:
The patent changes the fundamental parameters from full 3D electromagnetic field distributions to simplified 2D cross-sectional coupling coefficients. This parameter transformation maintains the essential physics of electromagnetic coupling while dramatically reducing computational requirements, as the coefficients can be extracted from 2D planes and reused for multiple prediction scenarios.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for more accurate and efficient prediction of SSO noise, reducing computational demands and enhancing confidence in noise prediction models, effectively addressing the limitations of existing methods.
Implementation Method 1
A method using superposition and pre-recorded noise waveforms to predict SSO noise by selecting victim and aggressor signal paths, recording electrical disturbances, and aggregating them to predict output noise
Data Source
AI summary
A method for the prediction of simultaneous switching output (SSO) noise that may be generated by one or more signal conduction paths within an electrical system. Electrical disturbance waveforms are first recorded for each signal conduction path that may be affected by the electrical disturbances. Next, principles of superposition are utilized to coherently combine each of the electrical disturbance waveforms in the time domain to generate the predicted SSO noise waveform that is imposed upon the affected signal conduction path. The electrical disturbance waveforms may be produced either by using bench measurements performed on an actual integrated circuit, by simulation, or by a combination of simulation and bench measurements.