Memory Device Identification via Physical Unclonable Function
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In memory systems, the inherent variations in fabrication processes lead to unpredictable results, making it challenging to identify memory devices reliably, as existing methods lack a robust mechanism for generating unique and unpredictable identifiers.
Innovation Solution
A system and method that utilize physical unclonable functions (PUFs) by performing raw read operations on memory blocks, sorting pages into low and high groups based on the average number of ones, generating unordered and ordered page pairs, and creating a sequence for identifying the block by comparing the average number of ones in each pair, thereby generating unique and unpredictable identifiers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional identification methods are used for memory devices, then the identification process is simple, but the reliability and uniqueness of the identifier is insufficient due to fabrication variations
Solution Approach 1:
The memory device automatically generates its own unique identifier by utilizing intrinsic fabrication variations within its own structure. The controller performs raw read operations on memory blocks and processes the data to create a PUF-based identifier, making the device self-identifying without requiring external identification components or complex external verification systems
Solution Approach 2:
The invention changes the identification approach from using manufactured identification tags to using electrical parameters (read voltages, threshold voltages) of memory blocks that naturally vary due to fabrication processes. By measuring and comparing these electrical parameters across multiple blocks, the system generates unique identifiers based on physical variations rather than artificial labels
2Reliability
If physical unclonable functions are used to generate unique identifiers, then the uniqueness and security are improved, but the complexity of the identification process increases
Solution Approach 1:
The identification process is segmented into distinct operational phases: selecting multiple memory blocks, performing raw read operations on each block, measuring threshold voltages, comparing voltage differences, and generating the final identifier. This segmentation allows each step to be independently optimized and executed, managing complexity through structured progression rather than monolithic processing
Solution Approach 2:
The controller acts as an intermediary that mediates between the physical memory blocks and the identification system. It performs the intermediate steps of reading raw data, calculating threshold voltages, comparing differences, and synthesizing the final unique identifier, thereby bridging the gap between physical variations and digital identification without requiring direct complex interaction between memory cells and external systems
3Measurement precision
If multiple raw read operations are performed on memory blocks, then the accuracy of threshold voltage measurement is improved, but the time consumption and operational complexity increases
Solution Approach 1:
The system performs preliminary raw read operations on memory blocks before final identifier generation. By conducting multiple read operations in advance and calculating average threshold voltages, the system establishes accurate baseline measurements that can be used for reliable identification while minimizing the time required during actual identification operations
Solution Approach 2:
The invention performs more read operations than strictly necessary for basic identification (excessive action) to ensure high measurement precision and reliability. By reading each block multiple times and averaging the results, the system compensates for random variations and noise, achieving robust threshold voltage measurements that withstand environmental variations and aging effects
Data Source
AI summary
A system which identifies a memory device using a physical unclonable function. The system performs raw read operations on every page of a block; sorts the pages into low and high groups using an average number of ones based on the raw read operations; generates unordered page pairs by sequentially selecting a first page from the low group and a second page from the high group; generates ordered page pairs by selectively converting an order of pages in each pair of the unordered page pairs; and generates a sequence for identifying the selected block based on comparing the average number of ones for pages in each ordered page pair.