This post is also available in: עברית (Hebrew)
As hacking techniques are constantly improving, scientists have been trying to find new ways and authentication methods to secure computers. Researchers in South Korea have developed a technique for randomly generating miniature mazes that could serve as identifying fingerprints for computer security. Their new research suggests that micro-mazes might one day serve as digital fingerprints for computer security.
Hyung Bae and colleagues from Seoul National University, South Korea have developed a new technology for creating specific mini-mazes by wrinkling a sheet of silica-based substrate.
According to cosmosmagazine.com, earlier research had used external forces and geometrical constraints to induce wrinkling in a smooth sheet of substrate, which resulted in the formation of ridges that move in various directions to create a maze. However, past wrinkling methods have created mini-mazes with little specificity due to a lack of control mechanisms.
Bae’s team have now developed an “elaborate wrinkle control process” to enhance the specificity of the mini-maze. The process involved creating a “photomask” – an opaque plate containing holes for light to pass through in a specific pattern. They describe their research in a paper in Science Advances.
UV light was then used to imprint exact geometric designs onto the plate in horizontal and vertical orientations. This was used as a guide for the wrinkling process by being placed underneath the silica platform. This meant that when wrinkling occurred the ridges could only move in specific, pre-ordained directions.
“Even though we have some control of the length scale or periodicity, we don’t know exactly what kind of pattern will appear when we first do this process,” says Sung Hoon Kang, assistant professor of mechanical engineering at Johns Hopkins University. According to discovermagazine.com, the virtue of this method is that order and randomness can coexist.
The intrinsic randomness of the process has always had interesting security applications—think tiny fingerprints or barcodes used to tag a device. This ‘guided randomness’ could be used as a unique identifier, but simultaneously provides enough control to encode some information into the pattern: a unique, but not arbitrary, identifier.
The innovation should allow for future mini-mazes to be structurally more precise for any application they are being prepared for. The authors believe that the mazes could be useful in computer security by serving as what is called “physical unclonable functions” – i.e. unique physical patterns that serve to undergird encryption and identification, in much the same way that fingerprints can uniquely identify people.