• The KIPS Transactions:PartC
• /
• v.12C no.1 s.97
• /
• pp.19-28
• /
• 2005

Recently, it has been sharply increased the interests to detect the network traffic anomalies to help protect the computer network from unknown attacks. In this paper, we propose a new anomaly detection scheme using the simple linear regression analysis for the exported LetFlow data, such as bits per second and flows per second, from a border router at a campus network. In order to verify the proposed scheme, we apply it to a real campus network and compare the results with the Holt-Winters seasonal algorithm. In particular, we integrate it into the RRDtooi for detecting the anomalies in real time.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.31 no.6
• /
• pp.1215-1225
• /
• 2021

Although Artificial Intelligence (AI) techniques have shown impressive performance in various fields, they are vulnerable to adversarial examples which induce misclassification by adding human-imperceptible perturbations to the input. Previous studies to defend the adversarial examples can be classified into three categories: (1) model retraining methods; (2) input transformation methods; and (3) adversarial examples detection methods. However, even though the defense methods against adversarial examples have constantly been proposed, there is no research to classify the type of adversarial attack. In this paper, we proposed an adversarial attack family classification method based on dimensionality reduction and clustering. Specifically, after extracting adversarial perturbation from adversarial example, we performed Linear Discriminant Analysis (LDA) to reduce the dimensionality of adversarial perturbation and performed K-means algorithm to classify the type of adversarial attack family. From the experimental results using MNIST dataset and CIFAR-10 dataset, we show that the proposed method can efficiently classify five tyeps of adversarial attack(FGSM, BIM, PGD, DeepFool, C&W). We also show that the proposed method provides good classification performance even in a situation where the legitimate input to the adversarial example is unknown.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.41 no.6 s.324
• /
• pp.17-24
• /
• 2004

This paper proposes an efficient technique to protect e-mail server from DDoS attack using the CBQ (Class Based Queuing) algorithm The proposed method classifies incoming trafic to an e-mail server into three classes: 'more important mail traffic', 'less important traffic' and 'unknown traffic' and assigns bandwidths differently to the traffics. By differentiating the bandwidths of classes, normal mail traffic may flow even under DDoS attack in the proposed technique. The proposed technique is implemented on an embedded system which hires a switching processor with the WFHBD(Weighted Fair Hashed Bandwidth Distribution) engine that has been known as an efficient algorithm to distribute a given bandwidth to multiple sources, and it is verified that it can be an efficient way to protect e-mail server from DDoS attack.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37C no.10
• /
• pp.949-964
• /
• 2012

Our PC have been under constant threat of malicious codes and viruses today. As many new ways of cyber attacks are being developed, such as zero-day-attack, nobody's PC is guaranteed to be safe from the attacks. In case where a user uses the existing verification protocol on a unsecured PC, the user's verification information may well be threatened by sniffing or man-in-the-middle attack. Especially, deadly attacks like memory hacking would give hard time for users to even recognize any symptom of virus infection. Therefore, this paper designs secured QR-Login user verification protocol for smart devices that are ready to communicate with QR-Code and proposes a way to keep critical data safe when using the internet. This way, user would be able to safeguard his/her critical data even when under attack by unknown attacks and safely carry out extremely sensitive task, like financial trading, on the device.

• Convergence Security Journal
• /
• v.18 no.4
• /
• pp.89-93
• /
• 2018

As terrorist organizations like ISIL and Al-Qaeda attack multiple-use facilities such as subways and department stores that many unknown citizens are using, the fear and anxiety are expanding. If international terrorism occurs in Korea, it may attack multiple-use facilities. Therefore, we need to have countermeasures against the attack. However, as the concept of multiple-use facilities has multi-faceted meanings and the scope is wide, there is a limitation in the preparation for international terrorism. In this study, it was tried to find the solution through reviewing the multiple-use facilities related laws. First of all, as the multiple-use facilities related laws were legislated by government organizations according to the necessities, the concept and scope are very inclusive. Therefore, it is possible to have controversies on the scope of the review related to anti-terrorism. Next, the department responsible for the international terrorism is not clear, which can make issue of rights and responsibility. To overcome such limitations, this study tried to set the concept and scope of multiple-use facilities and discussed the issue of the responsible department.

• International Journal of Aerospace System Engineering
• /
• v.7 no.1
• /
• pp.21-27
• /
• 2020

Quasi steady stall is a phenomenon to characterize the aerodynamic behavior of aircraft at high angle of attack region. Generally, it is exercised from a steady state level flight to stall and its recovery to the initial flight in a calm weather. For a theoretical study, such maneuver is demonstrated in the form of aerodynamic model which consists of aircraft's stability and control derivatives. The current research paper is focused on the appropriate selection of aerodynamic model for the maneuver and estimation of the unknown model coefficients using least-square method. The statistical accuracy of the estimated parameters is presented in terms of standard deviations. Finally, the validation has been presented by comparing the measured data to the simulated data from different models.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.04a
• /
• pp.359-361
• /
• 2003

기존의 오용 기반 침입탐지 시스템에서는 변형되거나 새로운 해킹 방법에 대한 지속적인 탐지패턴을 지원해 주어야 하는 단점이 있다. 이러한 변형되거나 알려지지 않은 공격에 대한 탐지는 비정상행위 탐지 방법으로 본 논문에서는 컴퓨터 면역시스템의 부정 및 긍정선택 방법과 가중치의 특성을 이용하였다. 즉, 알리진 공격으로부터 특성을 추출하여 알려지지 않은 공격에 대응할 수 있도록 특성을 변경하는 방법을 사용하였다. 이러한 방법으로 공격 특성을 추출하고 특성 추출에 사용하지 않은 다른 공격에 대한 탐지를 실험한 결과 u2r 공격인 buffer overflow 공격과 race condition 공격에 대하여 정확한 탐지가 이루어짐을 보였다.

• Proceedings of the Korea Institutes of Information Security and Cryptology Conference
• /
• 2003.12a
• /
• pp.141-145
• /
• 2003

We propose the model of Software Rejuvenation methodology, which is applicable for survivability. Software rejuvenation is a proactive fault management technique and being used in fault tolerant systems as a cost effective technique for dealing with software faults. Survivability focuses on delivery of essential services and preservation of essential assets, even systems are penetrated and compromised. Thus, our objective is to detect the intrusions in a real time and survive in face of such attacks. As we deterrent against an attack in a system level, the Intrusion tolerance could be maximized at the target environment. We address the optimal time to execute ad hoc software rejuvenation and we compute it by using the semi Markov process. This is one way that could be really frustrated and deterred the attacks, as the attacker can't make their progress. This Software Rejuvenation method can be very effective under the assumption of unknown attacks. In this paper, we compute the optimum time to perform an ad hoc Software Rejuvenation through intrusions.

• The KIPS Transactions:PartC
• /
• v.10C no.5
• /
• pp.525-532
• /
• 2003

This paper describes intrusion detection rule management using mobile agents. Intrusion detection can be divided into anomaly detection and misuse detection. Misuse detection is best suited for reliably detecting known use patterns. Misuse detection systems can detect many or all known attack patterns, but they are of little use for as yet unknown attack methods. Therefore, the introduction of mobile agents to provide computational security by constantly moving around the Internet and propagating rules is presented as a solution to misuse detection. This work presents a new approach for detecting intrusions, in which mobile agent mechanisms are used for security rules propagation. To evaluate the proposed approach, we compared the workload data between a rules propagation method using a mobile agent and a conventional method. Also, we simulated a rules management using NS-2 (Network Simulator) with respect to time.

• Journal of the Korea Computer Industry Society
• /
• v.5 no.8
• /
• pp.781-790
• /
• 2004

This paper describes intrusion detection rule mangement using mobile agents. Intrusion detection can be divided into anomaly detection and misuse detection. Misuse detection is best suited for reliably detecting known use patterns. Misuse detection systems can detect many or all known attack patterns, but they are of little use for as yet unknown attack methods. Therefore, the introduction of mobile agents to provide computational security by constantly moving around the Internet and propagating rules is presented as a solution to misuse detection. This work presents a new approach for detecting intrusions, in which mobile agent mechanisms are used for security rules propagation. To evaluate the proposed appraoch, we compared the workload data between a rules propagation method using a mobile agent and a conventional method. Also, we simulated a rules management using NS-2(Network Simulator) with respect to time.