An optical encryption technique developed by researchers at the Foundation for Research and Technology Hellas (FORTH) and the University of Crete in Greece is claimed to provide an exceptionally high level of security.
According to Optica, the system decodes the complex spatial information in the scrambled images by retrieving intricately jumbled information from a hologram using trained neural networks.
“From rapidly evolving digital currencies to governance, healthcare, communications and social networks, the demand for robust protection systems to combat digital fraud continues to grow," stated project leader Stelios Tzortzakis.
"Our new system achieves an exceptional level of encryption by utilizing a neural network to generate the decryption key, which can only be created by the owner of the encryption system.”
Optical encryption secures data at the network's optical transport level, avoiding slowing down the overall system with additional hardware at the non-optical levels. This strategy may make it easier to establish authentication procedures at both ends of the transfer to verify data integrity.
The researchers investigated whether ultrashort laser filaments travelling in a highly nonlinear and turbulent medium might transfer optical information, such as a target's shape, that had been encoded in holograms of those shapes. The researchers claim that this renders the original data totally scrambled and unretrievable by any physical modelling or experimental method.
Data scrambled by passage via ethanol liquid
A femtosecond laser was used in the experimental setup to pass through a prepared hologram and into a cuvette that contained liquid ethanol. A CCD sensor recorded the optical data, which appeared as a highly scrambled and disorganised image due to laser filamentation and thermally generated turbulence in the liquid.
"The challenge was figuring out how to decrypt the information," said Tzortzakis. “We came up with the idea of training neural networks to recognize the incredibly fine details of the scrambled light patterns. By creating billions of complex connections, or synapses, within the neural networks, we were able to reconstruct the original light beam shapes.”
In trials, the method was used to encrypt and decrypt thousands of handwritten digits and reference forms. After optimising the experimental approach and training the neural network, the encoded images were properly retrieved 90 to 95 percent of the time, with additional improvements potential with more thorough neural network training.
The team is now working on ways to use a less expensive and bulkier laser system, as a necessary step towards commercialising the approach for a variety of potential industrial encryption uses.
“Our study provides a strong foundation for many applications, especially cryptography and secure wireless optical communication, paving the way for next-generation telecommunication technologies," concluded Tzortzakis.
