Loading...
Please wait, while we are loading the content...
Similar Documents
A2-MpixelCMOSImageSensorwithDevice Authentication and Encryption Key Generation based on Physically Unclonable Function
| Content Provider | Semantic Scholar |
|---|---|
| Author | Okura, Shunsuke Ishiki, Ryota Takano, Syohei Shirahata, Masayoshi Kubota, Takaya Shiozaki, Mitsuru Ishikawa, Ken-Ichiro Takayanagi, Isao Fujino, T. |
| Copyright Year | 2019 |
| Abstract | To make the Internet of Things (IoT) a success, information security will have to be guaranteed. To achieve high enough information security, data confidentiality, data integrity, and device authentication are required. For such functions, the Physically Unclonable Function (PUF) [1]–[5] serves as a unique identifier (ID) and key for a device, based on physical variations caused during the manufacturing process. The strong dependence on the internal parameters makes a PUF a highly tamper-evident ID and key storage without non-volatile memory (NVM). Therefore the PUF can provide security that starts at the data source to prevent attackers from exploiting sensor networks. For the image information security, a 2 Mpixel 12 bit CMOS image sensor with a PUF (CIS-PUF) is proposed [6], in which the pixel-to-pixel fixed pattern noise (PPFPN) is utilized as a PUF ID of each device. The CIS-PUF based device authentication is realized by a challenge-response (CR) authentication, whose scheme consists of two phases, enrollment and verification. During the enrollment phase, the whole PUF ID bits derived from the pixel array are recorded by the verifier. During the verification phase, the verifier issues a challenge that is a randomly selected pixel address. The CIS must respond with the one string of PUF ID which fits the challenge the verifier issued. The verifier issues a different challenge each time, and thus knowing a previous correct response is of no use. The PUF can moreover be used to generate keys for cryptographic purposes such as data confidentiality and integrity, effectively binding the key to the hardware. The secret key initially generated is recorded by a host device. The CISPUF regenerates the key on demand to encrypt the image data, in which the key must be 100% recovered for the decryption by the host. In order to generate a cryptographic key removing noise present in the PUF response measurement, post-processing is required. For CIS-PUF, a dynamic softdecision error correction is proposed which realizes high error correction capability with small circuit overhead [7]. In this paper, evaluation results of the CR authentication are (a) Chip overview |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://www.imagesensors.org/Past%20Workshops/2019%20Workshop/2019%20Papers/P19.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
