New Class of Ciphers Using Hardware Entropy Source

TL;DR

The paper presents BARN, a novel symmetric cipher that hides a message within a hardware TRNG bitstream by inserting message bits at positions controlled by a secret key. The embedding relies on a simple, explicit rule that computes insertion indices using the key length $\kappa$ and a key sequence $K$, with decoding possible once the key is known; keys are generated from TRNG output by encoding digits in various bases, trading off ciphertext expansion against permutation space. Security is analyzed primarily as a brute-force search problem over the key space, with no arithmetic cryptanalytic attacks applicable, and the authors quantify the enormous search spaces for different base representations and key lengths. The work discusses practical deployment scenarios, including IoT and chip-scale TRNGs, and highlights throughput considerations (e.g., modest rates suffice for short messages, while high-throughput streams require faster TRNGs), positioning BARN as a lightweight, hardware-entropy-based approach suitable for niche cryptographic applications. The study thus suggests a niche where hardware randomness can enable low-complexity secure encoding with substantial brute-force resistance, particularly for systems with constrained computation and where silicon-integrated TRNGs are feasible.

Abstract

We present a novel, computationally simple method of hiding any message in the stream of random bits by using a secret key. The method is called Bury Among Random Numbers (BARN). A stream of random bits is produced by extracting the entropy of a physical process in a hardware-based true random number generator (TRNG). The process of placing bits of a message into the stream of random bits is governed by the number of random bits skipped between subsequent insertions. The set of numbers that correspond to the steps of BARN is derived from a random number also provided by TRNG. Hence BARN cipher does not depend on any arithmetic function. We propose an effective method of computing random keys from a given number of random bits. We estimate the number of permutations that need to be tested during a brute-force attack on the new cipher for various key lengths. Some practical applications for the new class of symmetrical ciphers are discussed.

Figures (4)

• Figure 1: Stream $P$ of random bits.
• Figure 2: Message $M$.
• Figure 3: Set of random bits $P$ with embedded message $M$ using key $K=\{1,3,2,1\}$.
• Figure 4: Decryption with guessed key $K=\{2,1,2,1\}$.