Design of Short Blocklength Wiretap Channel Codes: Deep Learning and Cryptography Working Hand in Hand

We design short blocklength codes for the Gaussian wiretap channel under information-theoretic security guarantees. Our approach consists in decoupling the reliability and secrecy constraints in our code design. Specifically, we handle the reliability constraint via an autoencoder, and handle the secrecy constraint via hash functions. For blocklengths smaller than 16, we evaluate through simulations the probability of error at the legitimate receiver and the leakage at the eavesdropper of our code construction. This leakage is defined as the mutual information between the confidential message and the eavesdropper’s channel observations, and is empirically measured via a recent mutual information neural estimator. Simulation results provide examples of codes with positive rates that achieve a leakage inferior to one percent of the message length.