Image Sensor FPN for Unique Device Identity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing device identification methods, such as IMEI and MAC addresses, are vulnerable to duplication and forgery due to software modifications, making it difficult to guarantee the uniqueness of a device's identity.
Innovation Solution
A method utilizing the fixed pattern noise (FPN) of a CMOS image sensor to generate a unique signature for device identification, which involves reading and processing column FPN information from the image sensor, downsampling, and comparing it with stored FPN information to perform certification and encryption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional identification methods (IMEI, MAC address) are used for device identification, then device identification is simple and convenient, but the identification can be easily duplicated, forged, or modified making uniqueness cannot be guaranteed
Solution Approach 1:
The patent replaces software-based identification systems (IMEI, MAC address) with a hardware-based physical unclonable function derived from the inherent fixed pattern noise characteristics of the image sensor. This substitution leverages the physical properties of the sensor hardware to generate a unique, unclonable identification signature, thereby ensuring uniqueness while maintaining system simplicity.
Solution Approach 2:
The patent converts the previously harmful or useless fixed pattern noise in image sensors into a beneficial resource for secure device identification. By utilizing these inherent noise characteristics, the system creates a unique fingerprint for each device that cannot be replicated, transforming a defect into a security feature.
2Reliability
If FPN information from multiple rows is processed to improve identification reliability, then identification accuracy increases, but processing time and computational complexity increase
Solution Approach 1:
The patent extracts only the essential FPN characteristics from selected rows of the image sensor data, rather than processing the entire dataset. By identifying and utilizing key features that sufficiently distinguish devices, the system achieves high identification accuracy while minimizing processing time and computational resources required.
Solution Approach 2:
The patent processes FPN information from a subset of rows rather than all rows, demonstrating that partial processing can achieve sufficient identification reliability. This selective approach balances accuracy requirements with processing efficiency, avoiding unnecessary computational overhead.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A mobile device method for certifying a mobile device includes: generating first fixed pattern noise (FPN) information based on column FPN of an image sensor included in the mobile device; and controlling the mobile device to perform a certification by using the first FPN information.