SRAM Random Number Generation via Multi-Voltage Cell Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Memory devices with imbalanced memory cells produce inconsistent Physically Unclonable Function (PUF) values due to balanced cells and varying minimum operating voltages, leading to unacceptable authentication inconsistencies.
Innovation Solution
An information handling system performs multi-voltage tests on memory cells to identify passing, randomly failing, and failing cells, generating a unique PUF value or random number based on their addresses and state values, leveraging manufacturing imperfections to create device-unique identifiers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multi-voltage tests are performed on memory cells to identify consistent cells, then PUF value consistency is improved, but device complexity increases
Solution Approach 1:
The memory device is divided into multiple memory cells, each tested independently at multiple voltage levels. This segmentation allows the system to identify consistent cells across voltage transitions while maintaining manageable test complexity through parallel processing of individual cells.
Solution Approach 2:
The system performs preliminary multi-voltage testing during manufacturing or initialization to identify cells that maintain consistent states across voltage transitions. This preliminary identification creates a mapped set of consistent cells that can be used for PUF generation without requiring repeated complex testing during operation.
2Adaptability or versatility
If memory cells with varying minimum operating voltages are used, then device uniqueness is improved, but PUF value consistency deteriorates
Solution Approach 1:
The system identifies cells with specific local characteristics - namely, cells that maintain consistent logical states across voltage transitions despite varying minimum operating voltages. By selecting only those cells with the desired local quality of voltage insensitivity, the system achieves both device uniqueness and PUF consistency.
Solution Approach 2:
The system changes the operating parameter (voltage level) across multiple test conditions to identify cells whose state remains invariant. By observing cell behavior across a range of voltage parameters, the system distinguishes between cells that are sensitive to voltage changes and those that maintain stable states, selecting the latter for PUF generation.
3Ease of manufacture
If balanced memory cells are included in PUF generation, then manufacturing simplicity is improved, but authentication reliability deteriorates
Solution Approach 1:
The system extracts and excludes balanced or inconsistent memory cells from the PUF generation process by identifying cells that fail to maintain consistent states across voltage transitions. Only cells that demonstrate stable, unpredictable behavior consistent with manufacturing variations are selected, removing problematic balanced cells that would compromise authentication reliability.
Data Source
AI summary
An approach is provided in which an information handling system performs multiple tests on a memory device using different supply voltage levels. The information handling system identifies a set of memory cells in the memory that produce a same set of results during each of the memory tests at the different supply voltage levels, and generates a random number based on a set of data values collected from the set of memory cells. In turn, the information handling system uses the random number generator in one or more processes executed by the information handling system.