Telecommunications Forum (TELFOR)

Security Verification of Artificial Neural Networks Used to Error Correction in Quantum Cryptography

Marcin Niemiec, Miralem Mehić, Miroslav Vozn̆ák

Published Nov 1, 2018• 24 citations

Abstract

Error correction in quantum cryptography based on artificial neural networks is a new and promising solution. In this paper the security verification of this method is discussed and results of many simulations with different parameters are presented. The test scenarios assumed partially synchronized neural networks, typical for error rates in quantum cryptography. The results were also compared with scenarios based on the neural networks with random chosen weights to show the difficulty of passive attacks.

DOI: 10.1109/telfor.2018.8612006

Related papers

Secure Synchronization of Artificial Neural Networks Used to Correct Errors in Quantum Cryptography

Marcin Niemiec, Tymoteusz Widlarz, Miralem Mehić

Quantum cryptography can provide a very high level of data security. However, a big challenge of this technique is errors in quantum channels. Therefore, error correction methods must be applied in real implementations. An example is error correction based on artificial neural networks. This paper considers the practical aspects of this recently proposed method and analyzes elements which influence security and efficiency. The synchronization process based on mutual learning processes is analyzed in detail. The results allowed us to determine the impact of various parameters. Additionally, the paper describes the recommended number of iterations for different structures of artificial neural networks and various error rates. All this aims to support users in choosing a suitable configuration of neural networks used to correct errors in a secure and efficient way.

arXiv:2301.11440International Conference on CommunicationsMay 28, 20232 citations

Error correction in quantum cryptography based on artificial neural networks

Marcin Niemiec

Intensive work on quantum computing has increased interest in quantum cryptography in recent years. Although this technique is characterized by a very high level of security, there are still challenges that limit the widespread use of quantum key distribution. One of the most important problems remains secure and effective mechanisms for the key distillation process. This article presents a new idea for a key reconciliation method in quantum cryptography. This proposal assumes the use of mutual synchronization of artificial neural networks to correct errors occurring during transmission in the quantum channel. Users can build neural networks based on their own string of bits. The typical value of the quantum bit error rate does not exceed a few percent; therefore, the strings are similar and also users’ neural networks are very similar at the beginning of the learning process. It has been shown that the synchronization process in the new solution is much faster than in the analogous scenario used in neural cryptography. This feature significantly increases the level of security because a potential eavesdropper cannot effectively synchronize their own artificial neural networks in order to obtain information about the key. Therefore, the key reconciliation based on the new idea can be a secure and efficient solution.

arXiv:1810.00957Quantum Information ProcessingApr 25, 201964 citations

Simulator of Quantum Bit Error Correction Process based on Artificial Neural Networks

Tymoteusz Widlarz, Marcin Niemiec

Quantum key distribution is a promising technology for ensuring secure communication in the era of quantum computing. However, its practical implementation faces challenges related to error correction due to the imperfections of quantum channels. Recently, artificial neural networks have shown significant potential for enhancing this process. This paper introduces a simulation tool for analyzing the security and efficiency of artificial neural networks in correcting errors within quantum key distribution systems. The tool explores various configurations, error levels, and synchronization dynamics, providing insights into the iteration requirements for legitimate parties compared to an attacker attempting to interfere. The simulator’s flexibility allows for the implementation of alternative learning algorithms and input types, offering a foundation for advancing neural network-based error correction in quantum cryptography. Additionally, the study highlights the impact of quantum bit error rates and error types on correction performance.

International Conference on Quantum Communications, Networking, and Computing (QCNC)Mar 31, 20250 citations

A New Error Correction Technique in Quantum Cryptography using Artificial Neural Networks

Gobinda Chandra Das, Malay Kule

Information Security is of paramount importance in digital communication in today’s highly networked world. Conventional information security methods face many future threats, one being the likelihood of large scale Quantum Computation. Quantum Cryptography is coming to the fore in the field of information security. It is all about using quantum physics to perform the cryptographic tasks as well as to break cryptographic systems. There are recent many advancements in the quantum cryptography field like Quantum Coin Flipping, Quantum Bit Commitment, and Quantum Key Distribution (QKD) networks. In this paper, a new concept for error correction and key reconciliation method is presented that uses Artificial Neural Network (ANN) with the help of Tree Parity Machine (TPM). In our work, we have considered a few cases of TPMs synchronization process with specific values of Quantum Bit Error Rate(QBER) and have performed some comparison among those. The experimental results observed into this paper that prove the correctness of the proposed work.

IEEE India ConferenceNov 24, 20225 citations

Quantum Error Mitigation With Artificial Neural Network

Changjun Kim, Daniel K. Park, June‐Koo Kevin Rhee

A quantum error mitigation technique based on machine learning is proposed, which learns how to adjust the probabilities estimated by measurement in the computational basis. Neural networks in two different designs are trained with random quantum circuits consisting of a set of one- and two-qubit unitary gates whose measurement statistics in the ideal (noiseless) and real (noisy) cases are known. Once the neural networks are trained, they infer the amount of probability adjustment to be made on the measurement obtained from executing an unseen quantum circuit to reduce the error. The proposed schemes are tested with two-, three-, five-, and seven-qubit quantum circuits of depth up to 20 by computer simulations with realistic error models and experiments using the IBM quantum cloud platform. In all test cases, the proposed mitigation technique reduces the error effectively. Our method can be used to improve the accuracy of noisy intermediate-scale quantum (NISQ) algorithms without relying on extensive error characterization or quantum error correction.

IEEE AccessJan 1, 202071 citations

Training artificial neural network to realize and mitigate general quantum error in real quantum devices

Chi Hang Chan

The field of quantum computing is quickly gaining attention from the society since the quantum supremacy claimed by Google in 2019. It can bring many applications such as many body system simulation and cryptography to a more advance level, makes it a multi-discipline subject. Same as the classical computers, error correction is essential in order to do any quantum computation, which is called quantum error correction(QEC). However, designing QEC codes require an error model to predict the error types, which is manufactured and may not be compatible for any situation. Also, since we are in the noisy intermediate-scale quantum(NISQ) era, QEC is hard to implement due to technical issues. In this work, I trained a neural network to mitigate the error instead of using QEC code to correct it. The network can identify the background noise and tells us what the uninterrupted states look like by reproducing them. Yet, the accuracy can be further improved by adopting different classification approach, which may also reduce the memory requirement so that it can deal with more qubits. The present work gives a concept on building a fast and neat neural network to mitigate general quantum errors, as well as to realize the background error, which may help us better understand the noise and describe them in a model.

Jan 1, 20220 citations