Neural Cryptography Research Articles

Cruciality of securing user data due to increasing Digital Learner traffic over the Internet using Adversarial Neural Cryptography

Cruciality of securing user data due to increasing Digital Learner traffic over the Internet using Adversarial Neural Cryptography

MAN[sbnd]C: A masked autoencoder neural cryptography based encryption scheme for CT scan images

Sharing medical images securely is very important towards keeping patients’ data confidential. In this paper we propose MANC: a Masked Autoencoder Neural Cryptography based encryption scheme for sharing medical images. The proposed technique builds upon recently proposed masked autoencoders. In the original paper, the masked autoencoders are used as scalable self-supervised learners for computer vision which reconstruct portions of originally patched images. Here, the facility to obfuscate portions of input image and the ability to reconstruct original images is used an encryption-decryption scheme. In the final form, masked autoencoders are combined with neural cryptography consisting of a tree parity machine and Shamir Scheme for secret image sharing. The proposed technique MANC helps to recover the loss in image due to noise during secret sharing of image.•Uses recently proposed masked autoencoders, originally designed as scalable self-supervised learners for computer vision, in an encryption-decryption setup.•Combines autoencoders with neural cryptography - the advantage our proposed approach offers over existing technique is that (i) Neural cryptography is a new type of public key cryptography that is not based on number theory, requires less computing time and memory and is non-deterministic in nature, (ii) masked auto-encoders provide additional level of obfuscation through their deep learning architecture.•The proposed scheme was evaluated on dataset consisting of CT scans made public by The Cancer Imaging Archive (TCIA). The proposed method produces better RMSE values between the input the encrypted image and comparable correlation values between the input and the output image with respect to the existing techniques.

Deep-Learning-based Cryptanalysis through Topic Modeling

Neural cryptography is a technique that uses neural networks for secure data encryption. Cryptoanalysis, on the other hand, deals with analyzing and decrypting ciphers, codes, and encrypted text without using a real key. Chosen-plaintext cryptanalysis is a subfield of cryptanalysis where both plain text and ciphertext are available and the goal is either to find the encryption technique, the encryption key, or both. This study addresses chosen plaintext cryptanalysis within public key cryptography, to categorize topics of encrypted text. Using a fixed encryption technique and key, the focus was placed on creating a framework that identifies the topic associated with ciphertext, using diverse plaintexts and their corresponding cipher texts. To our knowledge, this is the first time that chosen-plaintext cryptanalysis has been discussed in the context of topic modeling. The paper used deep learning techniques such as CNNs, GRUs, and LSTMs to process sequential data. The proposed framework achieved up to 67% precision, 99% recall, 80% F1-score, and 71% AUPR on a dataset, showcasing promising results and opening avenues for further research in this cryptanalysis subarea.

Synchronization of Tree Parity Machines Using Nonbinary Input Vectors.

Neural cryptography is the application of artificial neural networks (ANNs) in the subject of cryptography. The functionality of this solution is based on a tree parity machine (TPM). It uses ANNs to perform secure key exchange between network entities. This brief proposes improvements to the synchronization of two TPMs. The improvement is based on learning ANN using input vectors that have a wider range of values than binary ones. As a result, the duration of the synchronization process is reduced. Therefore, TPMs achieve common weights in a shorter time due to the reduction of necessary bit exchanges. This approach improves the security of neural cryptography.

Retracted: Intelligent CO2 Monitoring for Diagnosis of Sleep Apnea Using Neural Cryptography Techniques

Retracted: Intelligent CO₂ Monitoring for Diagnosis of Sleep Apnea Using Neural Cryptography Techniques

A novel medical steganography technique based on Adversarial Neural Cryptography and digital signature using least significant bit replacement

With recent advances in technology protecting sensitive healthcare data is challenging. Particularly, one of the most serious issues with medical information security is protecting of medical content, such as the privacy of patients. As medical information becomes more widely available, security measures must be established to protect confidentiality, integrity, and availability. Image steganography was recently proposed as an extra data protection mechanism for medical records. This paper describes a data-hiding approach for DICOM medical pictures. To ensure secrecy, we use Adversarial Neural Cryptography with SHA-256 (ANC-SHA-256) to encrypt and conceal the RGB patient picture within the medical image’s Region of Non-Interest (RONI). To ensure anonymity, we use ANC-SHA-256 to encrypt the RGB patient image before embedding. We employ a secure hash method with 256bit (SHA-256) to produce a digital signature from the information linked to the DICOM file to validate the authenticity and integrity of medical pictures. Many tests were conducted to assess visual quality using diverse medical datasets, including MRI, CT, X-ray, and ultrasound cover pictures. The LFW dataset was chosen as a patient hidden picture. The proposed method performs well in visual quality measures including the PSNR average of 67.55, the NCC average of 0.9959, the SSIM average of 0.9887, the UQI average of 0.9859, and the APE average of 3.83. It outperforms the most current techniques in these visual quality measures (PSNR, MSE, and SSIM) across six medical assessment categories. Furthermore, the proposed method offers great visual quality while being resilient to physical adjustments, histogram analysis, and other geometrical threats such as cropping, rotation, and scaling. Finally, it is particularly efficient in telemedicine applications with high achieving security with a ratio of 99% during remote transmission of Electronic Patient Records (EPR) over the Internet, which safeguards the patient’s privacy and data integrity.

α-Synchronization of a Class of Unbounded Delayed Inertial Cohen–Grossberg Neural Networks with Delayed Impulses

As an essential dynamic behavior, the synchronization of inertial Cohen–Grossberg neural networks (ICGNNs) has received considerable attention due to its successful applications in neural cryptography, public channel cryptography, security communications, and image encryption. In this article, the α-synchronization of a class of non-autonomous unbounded delayed inertial Cohen–Gossberg neural networks with delayed impulses is investigated. Firstly, several non-autonomous impulsive differential inequalities are established, where unbounded delays, delayed impulses, and time-variable coefficients are incorporated. Subsequently, based on the proposed impulsive differential inequalities and Lyapunov function approach, the feedback controllers are designed, and some criteria for α-synchronization are provided. Finally, the validity of the presented theoretical findings is demonstrated by two specific examples. It is shown that delayed impulses can be viewed as perturbations or stabilizing sources for non-autonomous ICGNNs.

Security of Neural Network-Based Key Agreement Protocol for Smart Grids

Recent developments in quantum computing pose a significant threat to the asymmetric cryptography currently in use. Neural cryptography offers a potential alternative that is resistant to attacks of known quantum computer algorithms. The considered solution is lightweight and computationally efficient. If a quantum computer algorithm were successfully implemented, it could expose IoT sensors and smart grid components to a wide range of attack vectors. Given the lightweight nature of neural cryptography and the potential risks, neural cryptography could have potential applications in both IoT sensors and smart grid systems. This paper evaluates one of the suggested enhancements: the use of integer-valued input vectors that accelerate the synchronization of the Tree Parity Machine. This enhancement introduces a new parameter M that indicates the minimum and maximum values of input vector elements. This study evaluates the nonbinary version of the mutual learning algorithm in a simulated insecure environment. The results indicate that, while the Nonbinary Tree Parity Machine may involve some trade-offs between security and synchronization time, the speed improvement is more substantial than the decrease in security. The impact of this enhancement is particularly significant for smaller adjustments to parameter M.

Neural Cryptography with Fog Computing Network for Health Monitoring Using IoMT

Sleep apnea syndrome (SAS) is a breathing disorder while a person is asleep. The traditional method for examining SAS is Polysomnography (PSG). The standard procedure of PSG requires complete overnight observation in a laboratory. PSG typically provides accurate results, but it is expensive and time consuming. However, for people with Sleep apnea (SA), available beds and laboratories are limited. Resultantly, it may produce inaccurate diagnosis. Thus, this paper proposes the Internet of Medical Things (IoMT) framework with a machine learning concept of fully connected neural network (FCNN) with k-nearest neighbor (k-NN) classifier. This paper describes smart monitoring of a patient’s sleeping habit and diagnosis of SA using FCNN-KNN+ average square error (ASE). For diagnosing SA, the Oxygen saturation (SpO2) sensor device is popularly used for monitoring the heart rate and blood oxygen level. This diagnosis information is securely stored in the IoMT fog computing network. Doctors can carefully monitor the SA patient remotely on the basis of sensor values, which are efficiently stored in the fog computing network. The proposed technique takes less than 0.2 s with an accuracy of 95%, which is higher than existing models.

Data protection based neural cryptography and deoxyribonucleic acid

The need to a robust and effective methods for secure data transferring makes the more credible. Two disciplines for data encryption presented in this paper: machine learning and deoxyribonucleic acid (DNA) to achieve the above goal and following common goals: prevent unauthorized access and eavesdropper. They used as powerful tool in cryptography. This paper grounded first on a two modified Hebbian neural network (MHNN) as a machine learning tool for message encryption in an unsupervised method. These two modified Hebbian neural nets classified as a: learning neural net (LNN) for generating optimal key ciphering and ciphering neural net CNN) for coding the plaintext using the LNN keys. The second granulation using DNA nucleated to increase data confusion and compression. Exploiting the DNA computing operations to upgrade data transmission security over the open nets. The results approved that the method is effective in protect the transferring data in a secure manner in less time

Neural Network Based Message Concealment Scheme

Abstract: Neural Cryptography is a new thread that integrates cryptography and neural networks for cryptanalysis and encryption applications. We show that Neural Networks can execute symmetric encryption in an adversarial context in this paper, and we build on the existing literature on the subject. Cryptography's purpose is to make it difficult to decipher a cypher and recreate the plain text without the associated key. Your messages are encrypted with excellent cryptography in such a way that brute force attacks against the algorithm or key are nearly impossible. Good cryptography is secure because it employs extremely long keys and encryption techniques that are resistant to various types of attack. The neural net application is the next step in the evolution of good cryptography. This paper discusses the use of neural networks in cryptography, including how to create neural networks that can be utilized in cryptography. Keywords: Cryptography key, encryption system, encryption algorithm, artificial neural network,chaos maps, logistic encryption.

Optimization of a Deep Learning Algorithm for Security Protection of Big Data from Video Images.

With the rapid development of communication technology, digital technology has been widely used in all walks of life. Nevertheless, with the wide dissemination of digital information, there are many security problems. Aiming at preventing privacy disclosure and ensuring the safe storage and sharing of image and video data in the cloud platform, the present work proposes an encryption algorithm against neural cryptography based on deep learning. Primarily, the image saliency detection algorithm is used to identify the significant target of the video image. According to the significant target, the important region and nonimportant region are divided adaptively, and the encrypted two regions are reorganized to obtain the final encrypted image. Then, after demonstrating how attackers conduct attacks to the network under the ciphertext attack mode, an improved encryption algorithm based on selective ciphertext attack is proposed to improve the existing encryption algorithm of the neural network. Besides, a secure encryption algorithm is obtained through detailed analysis and comparison of the security ability of the algorithm. The experimental results show that Bob's decryption error rate will decrease over time. The average classification error rate of Eve increases over time, but when Bob and Alice learn a secure encryption network structure, Eve's classification accuracy is not superior to random prediction. Chosen ciphertext attack-advantageous neural cryptography (CCA-ANC) has an encryption time of 14s and an average speed of 69mb/s, which has obvious advantages over other encryption algorithms. The self-learning secure encryption algorithm proposed here significantly improves the security of the password and ensures data security in the video image.

Post-COVID-19 Neural Cryptographic Onset of Homeopathy Psychiatry Medicine's Transmission in "New Normal Form"

In telepsychiatric treatments, the medical data of various types of patients' data are to be communicated in a well-protected way between the patient and the psychiatrists. Homeopathy drugs play a significant role in treating such psychiatric patients. The objective of this manuscript is to develop a strong session key with the help of neural cryptography. Neural cryptography is the extensive usage of the artificial neural networks for the cryptographic functions. E-medicines were encrypted here by the classical methods and the proposed session keys of 128 bits long. The length has been selected to 128 bits in order to have less memory space needed. Alzheimer's disease, Dementia, Insomnia, Stress, OCD, Brain restlessness, Depression, etc are the most clinical mental complications observed in this post COVID-19 situations. This is the era of digital health to deal the remote patients who are non-invasive in nature. In this manuscript, a session key has been proposed by the Key Distribution Centre (KDC) with the aid of artificial neural networks. Strong session keys were derived of 128 bits length which in turn used for AES 128 encryption. To avoid the Man-InThe-Middle attacks, it has been designed that the server machine shall accumulate all the needful symptoms of the psychiatric patients.

Intelligent CO2 Monitoring for Diagnosis of Sleep Apnea Using Neural Cryptography Techniques

In biomass wastage, carbon is one of the adsorbent materials. Biomass wastage contains complex materials, and pressure, various temperatures, and presence of various chemical components which are subjected to the adsorption of carbon are a tedious task, and it is used in the sustainable waste management system. While screening the biomass wastage management system, prediction of activated carbon’s quality and understanding of the mechanism of adsorption of [Formula: see text] are a complicated task. Many research works have been developed; the main issues are inaccurate and inefficient prediction of carbon available in the various feedstock of biomass wastage. To overcome these issues, this paper proposed gene expression programming (GEP) with [Formula: see text]-nearest neighbour (GEP-KNN). The key advantage of the proposed work shows excellent performance in the prediction of adsorbing carbon and accuracy. The accuracy of the GEP-KNN algorithm with different [Formula: see text] values produced the highest accuracy at [Formula: see text] and [Formula: see text] of 95.12% and 95.67%; the lowest accuracy is [Formula: see text] of 65.34%.

A neural cryptography approach for digital image security using Vigenère cipher and tree parity machine

As with other form of digital media, image can also contain private information that should be kept secret. Commonly, symmetric and asymmetric cryptosystem can be used to secure an image. Symmetric-key cryptography is not a best solution as the secret key has to be transmitted through private channel. Asymmetric-key cryptosystem, however, allows transmission of key through public channel, but requires high computational power and long computational time. This paper proposes a neural cryptography approach to protect image by using tree parity machine and Vigenère cipher. By using this approach, the secret key will never be transmitted. Moreover, the whole process does not require high computational power and long computational time as in public-key cryptosystem. To overcome the weakness of Vigenère cipher, change is made to the algorithm by multiplying the key value with y-axis of image. The effectiveness is validated by the value of pearson correlation that is close to 0. Lastly, the result of experiments show neural synchronization time depends on the size of network, initial weights and randomly generated input vector.

An FPGA-Based Performance Evaluation of Artificial Neural Network Architecture Algorithm for IoT

Nowadays, the high number of devices and applications connected to the Internet has generated a great amount of data being which makes privacy and protection a more challenging task. In addition, new technologies, such as the Internet of Things, incorporate many resource-constrained devices in the network. Reliable cryptography algorithms have to be employed to deal with this problem, which also needs to be efficiently implemented in small devices. There are several algorithms for this purpose, among them, neural cryptography. In this context, this work proposes the implementation of an artificial neural network architecture called tree parity machine (TPM) to perform the exchange of keys through the mutual learning of these networks. This method is not based on number theory, which makes it less computationally costly, and can be an alternative for embedded systems, which generally have several limitations in the processing capacity and resources used. In the area of embedded systems, FPGAs have gained more space, thanks to their reconfiguration capacity. Thus, different methods for implementing a TPM in FPGA were tested and analyzed, in order to optimize the following performance parameters, the response time, the maximum frequency of operation, and the consumed area of the FPGA considering logical elements, embedded multipliers, and registers. In addition, software implementation based on a multi-core CPU was used for comparison purposes. Experimental results demonstrated that the implementation of parallelism in FPGA for different blocks of the TPM weight matrix reached the best performance results. Thus, our proposal intends to develop an economic component in terms of resource consumption, however, maintaining the characteristic of high processing capacity. Therefore, the methodologies presented in this paper intends to be a useful reference to optimize future implementations in FPGA for cryptography applications.

Neural synchronization of optimal structure-based group of neural networks

In this paper, a neural synchronization of the optimal structure-based group of neural networks is proposed. Asymmetric cryptography is widely used to generate a key amongst two parties and to exchange the key through an insecure channel. However, since the methods that used this strategy, like RSA, have been compromised, new methods for producing a key that can offer security must be discovered. A new group of cryptography known as neural cryptography was created to solve this issue. The primary aim of neural cryptography is to produce a secret key over an unreliable medium. The optimal neural network architecture for creating and defining a secret key between the two authorized individuals is examined in this article. Furthermore, studies into the coordination of a group of neural networks are uncommon. For the design of the public key exchange protocol, synchronization of a cluster of neural networks with Three Layer Tree Parity Machine (TLTPM) is proposed. To calculate the synchronization time, steps taken, and the number of times the attacking neural network could replicate the actions of the two accepted networks, more than 15 million simulations were run. Various parametric experiments have been conducted on the proposed methodology. Simulations of the approach show that it is correct, according to the results of the paper.

Image encryption based on adversarial neural cryptography and SHA controlled chaos

To overcome the vulnerability to known-plaintext attack or chosen-plaintext attack of a linear image encryption system, a new approach for image encryption is proposed based on adversarial neural cryptography (ANC) combined with SHA-256 controlled chaotic systems. In this image encryption approach, the optimal network model is first obtained by training a generative adversarial network (GAN), and then the GAN model is used to achieve a noise-like intermediate image. Subsequently, the XOR operation based on a logistic map is performed on the intermediate image to obtain the final ciphertext. The intrinsic non-linearity of the neural network (NN) guarantees the ability of the proposed system to resist common attacks like known-plaintext attack or chosen-plaintext attack. The plaintext dependent SHA-256 controlled logistic map greatly improves the diffusion performance so that the encryption system can resist differential attacks. Numerical simulation results prove the reliability, effectiveness, and security of the proposed scheme.

Neural cryptography using optimal structure of neural networks

The asymmetric cryptography method is typically used to transfer the key via an insecure channel while creating a key between two parties. However, since the methods using this strategy, like RSA, are now breached, new strategies must be sought to generate a key that can provide security. To solve this issue, a new group of cryptography was created known as neural cryptography. The main objective of this neural cryptography is to create a secret key using an unsafe medium. This paper suggests an overview of the optimal configuration of the neural network that enables the generation and establishment of a secret key between the two approved entities. Synchronization of two neural networks with three hidden layers is proposed for the development of the public key exchange protocol. Over 15 million simulations were carried out to measure the synchronization time, the steps taken as well as the number of times the assaulting neural network can replicate the behavior of the two authorized networks. The proposed technique has been passed through different parametric tests. Simulations of the process show effectiveness in terms of cited results in the paper.

Neural Cryptography Based on Generalized Tree Parity Machine for Real-Life Systems

Traditional public key exchange protocols are based on algebraic number theory. In another perspective, neural cryptography, which is based on neural networks, has been emerging. It has been reported that two parties can exchange secret key pairs with the synchronization phenomenon in neural networks. Although there are various models of neural cryptography, called Tree Parity Machine (TPM), many of them are not suitable for practical use, considering efficiency and security. In this paper, we propose a Vector-Valued Tree Parity Machine (VVTPM), which is a generalized architecture of TPM models and can be more efficient and secure for real-life systems. In terms of efficiency and security, we show that the synchronization time of the VVTPM has the same order as the basic TPM model, and it can be more secure than previous results with the same synaptic depth.