Postquantum symmetric cryptography inspired by neural networks
Wojciech Węgrzynek, Paweł Topa · Annals of Computer Science and Information Systems · 2023
We introduce a novel approach to postquantum symmetric encryption that allows us to modify and continue to use any encryption scheme.By composing the encryption and decryption functions with the evaluation of arbitrarily wide neural networks we are able to verify that anyone performing these functions has access to at least a certain amount of memory.Since the number of qubits in quantum computers has been relatively slow-growing, this provides us security from the Grover's search attack, and any attack utilizing a similar oracle circuit.