• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2015.05a
• /
• pp.259-263
• /
• 2015

The vulnerabilities related to Flash player which is widely used in internet browsers and office programs are gradually increased. To detect Flash malwares, previous work focuses on predefined features of ActionScript. However above work cannot detect new/mutated Flash malwares, since predefined features could not cover the new patterns of new/mutated Flash mawares. To solve this problem, we propose a Flash malware detection method that uses machine learning to learn Flash Tag patterns and classify Flash by using machine learning.

• International journal of advanced smart convergence
• /
• v.8 no.2
• /
• pp.77-87
• /
• 2019

One of important concerns in information security is to control information flow. It is whether to protect confidential information from being leaked, or to protect trusted information from being tainted. In this paper, we present Piosk (Physical blockage of Information flow Kiosk) that addresses both the problems practically. Piosk can forestall and prevent the leakage of information, and defend inner tangible assets against a variety of malwares as well. When a visitor who carries a re-writable portable storage device, must insert the device into Piosk installed next to the security gate. Then, Piosk scans the device at the very moment, and detects & repairs malicious codes that might be exist. After that, Piosk writes the contents (including sanitized ones) on a new read-only portable device such as a compact disk. By doing so, the leakage of internal information through both insiders and outsiders can be prevented physically. We have designed and prototyped Piosk. The experimental verification of the Piosk prototype implementation reveals that, Piosk can accurately detect every malware at the same detection level as Virus Total and effectively prevent the leakage of internal information. In addition, we compare Piosk with the state-of-the-art methods and describe the special advantages of Piosk over existing methods.

• International Journal of Computer Science & Network Security
• /
• v.22 no.10
• /
• pp.237-245
• /
• 2022

The prominence of IoTs (Internet of Things) and exponential advancement of computer networks has resulted in massive essential applications. Recognizing various cyber-attacks or anomalies in networks and establishing effective intrusion recognition systems are becoming increasingly vital to current security. MLTs (Machine Learning Techniques) can be developed for such data-driven intelligent recognition systems. Researchers have employed a TFDNNs (Tensor Flow Deep Neural Networks) and DCNNs (Deep Convolution Neural Networks) to recognize pirated software and malwares efficiently. However, tuning the amount of neurons in multiple layers with activation functions leads to learning error rates, degrading classifier's reliability. HTFDNNs ( Hybrid tensor flow DNNs) and MRNs (Modified Residual Networks) or Resnet CNNs were presented to recognize software piracy and malwares. This study proposes HTFDNNs to identify stolen software starting with plagiarized source codes. This work uses Tokens and weights for filtering noises while focusing on token's for identifying source code thefts. DLTs (Deep learning techniques) are then used to detect plagiarized sources. Data from Google Code Jam is used for finding software piracy. MRNs visualize colour images for identifying harms in networks using IoTs. Malware samples of Maling dataset is used for tests in this work.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.25 no.4
• /
• pp.829-838
• /
• 2015

As the number of smartphone users have increased, many kinds of malwares have emerged. Unlike existing malwares, spyware can be installed normally after user authentication and agreement according to security policy. For this reason, it is not easy to catch spywares involving harmful functionalities to users by using existing malware detection system. Therefore, our paper focuses on study about detecting mainly wiretapping spywares among them by developing a new wiretapping detection model and application. Specifically, this study conducts to find out power consumption on each application and modular and network consumption to detect voice wiretapping so Open Source Project Power Tutor is used to do this. The risk assessment of wiretapping is measured by gathered all power consumption data from Open Source Project Power Tutor. In addition, developed application in our study can detect at-risk wiretapping spyware through collecting and analyzing data. After we install the application to the smartphone, we collect needed data and measure it.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.22 no.3
• /
• pp.667-677
• /
• 2012

Vulnerable web applications have been the primary method used by the attackers to spread their malware to a large number of victims. Such attacks commonly make use of malicious links to remotely execute a rather advanced malicious code. The attackers often deploy malwares that utilizes unknown vulnerabilities so-called "zero-day vulnerabilities." The existing computer vaccines are mostly signature-based and thus are effective only against known attack patterns, but not capable of detecting zero-days attacks. To mitigate such limitations of the current solutions, there have been a numerous works that takes a behavior-based approach to improve detection against unknown malwares. However, behavior-based solutions arbitrarily introduced a several limitations that made them unsuitable for real-life situations. This paper proposes an advanced web browser based malicious behavior detection system that solves the problems and limitations of the previous approaches.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.11
• /
• pp.4105-4121
• /
• 2021

Recently, malware classification based on Deep Neural Networks (DNN) has gained significant attention due to the rise in popularity of artificial intelligence (AI). DNN-based malware classifiers are a novel solution to combat never-before-seen malware families because this approach is able to classify malwares based on structural characteristics rather than requiring particular signatures like traditional malware classifiers. However, these DNN-based classifiers have been found to lack robustness against malwares that are carefully crafted to evade detection. These specially crafted pieces of malware are referred to as adversarial examples. We consider a clever adversary who has a thorough knowledge of DNN-based malware classifiers and will exploit it to generate a crafty malware to fool DNN-based classifiers. In this paper, we propose a DNN-based malware classifier that becomes resilient to these kinds of attacks by exploiting Generative Adversarial Network (GAN) based data augmentation. The experimental results show that the proposed scheme classifies malware, including AEs, with a false positive rate (FPR) of 3.0% and a balanced accuracy of 70.16%. These are respective 26.1% and 18.5% enhancements when compared to a traditional DNN-based classifier that does not exploit GAN.

• Journal of the Korea Society of Computer and Information
• /
• v.19 no.4
• /
• pp.81-90
• /
• 2014

As mobile devices have become widely used and developed, PC based malwares can be moving towards mobile-based units. In particular, mobile Botnet reuses powerful malicious behavior of PC-based Botnet or add new malicious techniques. Different from existing PC-based Botnet detection schemes, mobile Botnet detection schemes are generally host-based. It is because mobile Botnet has various attack vectors and it is difficult to inspect all the attack vector at the same time. In this paper, to overcome limitations of host-based scheme, we compare two network-based schemes which detect mobile Botnet by applying HMM and SVM techniques. Through the verification analysis under real Botnet attacks, we present detection rates and detection properties of two schemes.

• Journal of Information Technology Services
• /
• v.15 no.2
• /
• pp.157-167
• /
• 2016

Task environment for enterprises and public institutions are moving into cyberspace-based environment and structing the LTE wireless network. The applications "App" operated in the LTE wireless network are mostly being developed with Android-based. But Android-based malwares are surging and they are the potential DDoS attacks. DDoS attack is a major information security threat and a means of cyber attacks. DDoS attacks are difficult to detect in advance and to defense effectively. To this end, a DMZ is set up in front of a network infrastructure and a particular server for defensive information security. Because There is the proliferation of mobile devices and apps, and the activation of android diversify DDoS attack methods. a DMZ is a limit to detect and to protect against DDoS attacks. This paper proposes an information security method to detect and Protect DDoS attacks from the terminal phase using a Preemptive military strategy concept. and then DDoS attack detection and protection app is implemented and proved its effectiveness by reducing web service request and memory usage. DDoS attack detection and protecting will ensure the efficiency of the mobile network resources. This method is necessary for a continuous usage of a wireless network environment for the national security and disaster control.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.25 no.2
• /
• pp.321-330
• /
• 2015

There have been found many modified malwares which could avoid detection simply by replacing a sequence of characters or a part of code. Since the existing anti-virus program performs signature-based analysis, it is difficult to detect a malware which is slightly different from the well-known malware. This paper suggests a method of detecting modified malwares by extending a hash-value based code comparison. We generated hash values for individual functions and individual code blocks as well as the whole code, and thus use those values to find whether a pair of codes are similar in a certain degree. We also eliminated some numeric data such as constant and address before generating hash values to avoid incorrectness incurred from them. We found that the suggested method could effectively find inherent similarity between original malware and its derived ones.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2015.10a
• /
• pp.865-866
• /
• 2015

스마트폰은 주로 사용되고 있는 안드로이드 OS는 다양한 악성코드로 인해 금전적 피해, 데이터 유출 및 통제권한 상실 등과 같은 많은 피해를 당하고 있다. 침해 위협을 가중시키고 있는 모바일 악성코드 중 심각한 피해를 유발하는 커널 기반의 루팅(Rooting) 악성코드는 일반적인 탐지 방법으로는 찾아낼 수 없는 어려움이 있다. 본 논문에서는 커널 기반에서 동작하는 루팅(Rooting) 악성코드를 탐지하기 위한 방법을 제안한다. 스마트폰 어플리케이션이 실행될 때마다 생성되는 모든 프로세스의 UID를 확인하여 비정상적으로 사용자(User) 권한에서 관리자(Root) 권한으로 변환되는지를 확인하는 방법이다. 제안하는 방법을 활용하여 알려지지 않은 악성코드로 인한 안드로이드 OS의 피해를 최소화할 수 있을 것으로 기대된다.