• Smart Media Journal
• /
• v.12 no.2
• /
• pp.76-82
• /
• 2023

In the 4th industrial revolution, the blockchain that distributes and manages data through a P2P network is used as a new decentralized networking paradigm in various fields such as manufacturing, culture, and public service. However, with the advent of quantum computers, quantum algorithms that are able to break existing cryptosystems such as hash function, symmetric key, and public key cryptography have been introduced. Currently, because most major blockchain systems use an elliptic curve cryptography to generate signatures for transactions, they are insecure against the quantum adversary. For this reason, the research on the quantum-resistant blockchain that utilizes lattice-based cryptography for transaction signatures is needed. Therefore, in this paper, we propose a blind signature scheme for the blockchain in which the contents of the signature can be verified later, as well as signing by hiding the contents to be signed using lattice-based cryptography with the property of quantum resistance. In addition, we prove the security of the proposed scheme using a random oracle model.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.33 no.4
• /
• pp.687-697
• /
• 2023

Botnets, whose attack patterns are becoming more sophisticated and diverse, are recognized as one of the most serious cybersecurity threats today. This paper revisits the experimental results of botnet detection using autoencoder, a semi-supervised deep learning model, for UGR and CTU-13 data sets. To prepare the input vectors of autoencoder, we create data points by grouping the NetFlow records into sliding windows based on source IP address and aggregating them to form features. In particular, we discover a simple power-law; that is the number of data points that have some flow-degree is proportional to the number of NetFlow records aggregated in them. Moreover, we show that our power-law fits the real data very well resulting in correlation coefficients of 97% or higher. We also show that this power-law has an impact on the learning of autoencoder and, as a result, influences the performance of botnet detection. Furthermore, we evaluate the performance of autoencoder using the area under the Receiver Operating Characteristic (ROC) curve.

• Convergence Security Journal
• /
• v.23 no.1
• /
• pp.125-137
• /
• 2023

Companies that handle sensitive information, led by public institutions, establish separate networks for work and the Internet and protect important data through strong access control measures to prevent cyber attacks. Therefore, systems that involve the junction where the Intranet(internal LAN for work purposes only) and the Internet network are connected require the establishment of a safe security environment through both administrative and technical measures. Mobile Device Management(MDM) solutions to control mobile devices used by institutions are one such example. As this system operates by handling sensitive information such as mobile device information and user information on the Internet network, stringent security measures are required during operation. In this study, a model was proposed to manage sensitive information data processing in systems that must operate on the Internet network by managing it on the internal work network, and the function design and implementation were centered on an MDM solution based on a network interconnection solution.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.214-216
• /
• 2021

It proposes a plan to strengthen the limitations of existing corporate security systems based on Zero-Trust. With the advent of the era of the Fourth Industrial Revolution, the paradigm of security is also changing. As remote work becomes more active due to cloud computing and COVID-19, security issues arising from the changed IT environment are raised. At the same time, in the current situation where attack techniques are becoming intelligent and advanced, companies should further strengthen their current security systems by utilizing zero trust security. Zero-trust security increases security by monitoring all data communications based on the concept of doubting and trusting everything, and allowing strict authentication and minimal access to access requestors. Therefore, this paper introduces a zero trust security solution that strengthens the existing security system and presents the direction and validity that companies should introduce.

• Korean Journal of Social Welfare Studies
• /
• v.41 no.3
• /
• pp.57-79
• /
• 2010

This study set out to investigate internet experience effect on youth's delinquency. Specifically, internet experience means internet deviant behaviors and the frequency of the internet use including chatrooms, computer game, and pornography. Data came from Korea Youth Panel Survey 2007. Analysis methods hierarchical regression analysis. The major findings of this study are as follows. First, results showed that adolescents use computer every day about 2 hour 40 minutes for using internet such as chatrooms, computer game, and pornography. And 29.4% of adolescents reported internet deviance. Second, the internet use and the internet deviance influenced positively juvenile delinquency. Third, the relationship between internet use and juvenile delinquency was mediated by aggression and internet deviance. These results means that youth spend much time online every day for using internet, and engaged internet deviance. This online experience influence juvenile delinquency offline world. And limitations and implications of this study were discussed with respect to further studies.

• Korean Journal of Culture and Social Issue
• /
• v.18 no.3
• /
• pp.415-433
• /
• 2012

This study was conducted in order to explain the effect of driving behavior determinants such as drivers' personality and attitude that may induce risky driving behavior and to develop a valid method for discriminating risky drivers using the determinants. In the results of surveying 534 adult drivers, 5 driving behavior determinants (avoidance of problems, benefit/stimulus seeking, interpersonal anxiety, interpersonal anger, and aggression) were found to have a statistically significant effect on drivers' various risky driving behaviors. Using these factors, drivers were grouped according to risk levels (normal drivers, unintentionally risky drivers, and intentionally risky drivers). This result suggests that drivers' dangerous behavior level can be predicted using psychological factors such as their personality and attitude. Accordingly, if the driving behavior determinant model and the base score system used in this study are improved through further research, they are expected to be useful in predicting drivers' recklessness in advance, identifying problems, and providing differentiated safe driving education services based on the results.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.34 no.1
• /
• pp.21-29
• /
• 2024

With the development of ICT and network, security management of IT infrastructure that has grown in size is becoming very difficult. Many companies and public institutions are having difficulty managing system and network security. In addition, as the complexity of hardware and software grows, it is becoming almost impossible for a person to manage all security. Therefore, AI is essential for network security management. However, since it is very dangerous to operate an attack model in a real network environment, cybersecurity emulation research was conducted through reinforcement learning by implementing a real-life network environment. To this end, this study applied reinforcement learning to the network environment, and as the learning progressed, the agent accurately identified the vulnerability of the network. When a network vulnerability is detected through AI, automated customized response becomes possible.

• Convergence Security Journal
• /
• v.24 no.3
• /
• pp.3-11
• /
• 2024

Information security is crucial not only for protecting against external cyber-attacks but also for identifying and blocking internal data leakage risks in advance. To this end, many companies and institutions implement digital rights management(DRM) document security solutions, which encrypt files to prevent content access if leaked, and data loss prevention(DLP) solutions, which control devices such as USB ports on computing equipment to prevent data leaks. At a time when efforts to prevent internal data leaks are crucial, there is a growing need for control policies such as device control and the identification of information assets in standalone network environments, which could otherwise fall into unmanaged domains. In this study, we propose a Generation-Distribution-Application model for device control policies that are uniquely applied to standalone information assets that are not connected to internal networks. To achieve this, we developed an authentication technique linked with the asset management system, where information assets are automatically registered upon acquisition. This system allows for precise identification of information assets and enables flexible device control, and we have designed and implemented a system based on these principles.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.8
• /
• pp.103-112
• /
• 2024

This study was initiated with the aim of authenticating that inputs have not been tampered with without disclosing them in the case of computations where multiple inputs are entered by participants using the same key. In general, in the authentication stage, authentication is performed after the input value is disclosed, but we do not want to reveal the inputs until the end. This is a case of deviating from the traditional security model in which malicious participants exist in cryptography, but it is a malicious attack method that can actually occur enough. Privacy infringement or distortion of calculation results can occur due to malicious manipulation of input values. To prevent this, this study studied a method that can authenticate that the message is not a modified message without disclosing the message using the signature system, zero-knowledge proof, and commitment scheme. In particular, by modifying the ElGamal signature system and combining it with the commitment scheme and zero-knowledge proof, we designed and proved a verification protocol that the input data is not a modified data, and the efficiency was improved by applying batch verification between authentication.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.34 no.5
• /
• pp.833-840
• /
• 2024

Existing side-channel analysis studies have mostly been analyzed only on block ciphers without considering the operation mode. However, establishing a methodology of side-channel analysis on operation mode is necessary because information for performing analysis varies depending on that. This paper proposes a methodology of correlation power analysis (CPA) on an operation mode CBC in a software target. The first round SubBytes layer output is generally used as a sensitive hypothetical intermediate value of an encryption algorithm AES (advanced encryption standard); however, the adversary should acquire the plaintext and ciphertext to calculate the input of AES in CBC mode. We propose an intermediate value calculated only by ciphertext. Besides, the initial vector (IV) could be treated as closed information in practice, although it is theoretically not secret. The adversary cannot decrypt the first block of plaintext without IV even if he analyzes the secret key. We propose a deep learning-based IV analysis method in a non-profiled environment.