In next-generation network services such as the Internet of Things (IoT) and cyber-physical systems (CPS), which have been attracting attention in recent years, innumerable devices are connected to the network, resulting in unauthorized use, counterfeiting, and alteration of hardware. There is a concern about the threat caused by the Internet of Things.
In order to prevent such malicious attacks, hardware authentication (authentication judgment) is becoming more important, and a function (PUF: Physically Unclonable Function) that outputs a unique random value, also called a "fingerprint" of hardware, makes physical duplication difficult. Hardware authentication using a function) is considered to be a promising technology.However, since this technology utilizes minute differences in uncontrollable hardware, it is generally unstable and inefficient, and there is a strong demand for a stable and efficient method of using it.
Under these circumstances, the group of Research Institute of Electrical Communication, Tohoku University announced that it has developed a new method to authenticate the validity of hardware using PUF.With this new method, the bias of PUF output (ratio of 2 and 0) expressed in binary can be efficiently eliminated by a method based on rejection sampling, and the deterioration of safety due to the bias of PUF output can be prevented.In other words, even if it is a poor quality (low safety) "fingerprint", it is efficiently extracted and converted into a good quality (high safety) "fingerprint" and authenticated, so various "fingerprints" are cheap. It will be possible to use it with care.
In addition, with the world's highest conversion and authentication efficiency, we have achieved a significant reduction in mounting costs, which is less than half that of conventional devices, and sensors and ultra-compact information and communication equipment that were previously difficult to install hardware authentication. It is expected to be widely applied to.
In the future, the company plans to install this technology in an actual system and conduct demonstration experiments while advancing its application to new devices.
Paper information:[IACR Transactions on Cryptographic Hardware and Embedded Systems] Rejection Sampling Schemes for Extracting Uniform Distribution from Biased PUFs
