• 정보과학회 논문지
• /
• 제43권5호
• /
• pp.596-605
• /
• 2016

2009년 이후 정부 및 민간부문에서는 DDoS 방어체계 구축을 위해 수백억 원의 예산을 투입해 왔으며, 그 결과 많은 정부 및 민간분야에 DDoS 대응을 위한 전용장비가 설치되었다. 그러나 이러한 기관 역시 DDoS 공격 발생 시 성공적인 방어가 이루어지지 않는 경우가 많은데, 이는 DDoS 대응 장비가 특정 공격 행위에만 대응할 수 있는 시그니처 중심의 방어 구조를 따르고 있기 때문이다. 이에 비해 방어자원 관점의 정책적 대응방법을 통할 경우, 공격 기법과 상관없이 서비스 자원의 가용성 확인을 통하여 시스템 이상여부 및 공격 유형의 종류를 확인할 수 있으며, 공격에 대한 대응 정책 또한 손쉽게 도출할 수 있다. 본 고에서는 기존의 공격 행위 중심의 방어체계에서 벗어나 방어자 관점의 DDoS 탐지 기법을 소개하고, 이를 통해 정책기반 서비스거부공격 대응방안을 제시한다.

• International journal of advanced smart convergence
• /
• 제8권1호
• /
• pp.141-146
• /
• 2019

These days, there are many applications using neural networks as parts of their system. On the other hand, adversarial examples have become an important issue concerining the security of neural networks. A classifier in neural networks can be fooled and make it miss-classified by adversarial examples. There are many research to encounter adversarial examples by using denoising methods. Some of them using GAN (Generative Adversarial Network) in order to remove adversarial noise from input images. By producing an image from generator network that is close enough to the original clean image, the adversarial examples effects can be reduced. However, there is a chance when adversarial noise can survive the approximation process because it is not like a normal noise. In this chance, we propose a research that utilizes high-level representation in the classifier by combining GAN network with a trained U-Net network. This approach focuses on minimizing the loss function on high representation terms, in order to minimize the difference between the high representation level of the clean data and the approximated output of the noisy data in the training dataset. Furthermore, the generated output is checked whether it shows minimum error compared to true label or not. U-Net network is trained with true label to make sure the generated output gives minimum error in the end. At last, the remaining adversarial noise that still exist after low-level approximation can be removed with the U-Net, because of the minimization on high representation terms.

• 한국IT서비스학회지
• /
• 제15권2호
• /
• 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.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2005년도 가을 학술발표논문집 Vol.32 No.2 (1)
• /
• pp.214-216
• /
• 2005

We consider Wireless Sensor Network Security (WSN) and focus our attention to tolerate damage caused by an adversary who has compromised deployed sensor node to modify, block, or inject packets. We adopt a probabilistic secret sharing protocol where secrets shared between two sensor nodes are not exposed to any other nodes. Adapting to WSN characteristics, we incorporate these secrets to establish new pairwise key for node to node authentication and design multipath routing to multiple base stations to defend against HELLO flood attacks. We then analytically show that our defense mechanisms against HELLO flood attack can tolerate damage caused by an intruder.

• 한국컴퓨터정보학회논문지
• /
• 제22권10호
• /
• pp.83-92
• /
• 2017

Reconnaissance is performed gathering information from a series of scanning probes where the objective is to identify attributes of target hosts. Network reconnaissance of IP addresses and ports is prerequisite to various cyber attacks. In order to increase the attacker's workload and to break the attack kill chain, a few proactive techniques based on the network-based moving target defense (NMTD) paradigm, referred to as IP address mutation/randomization, have been presented. However, there are no commercial or trial systems deployed in real networks. In this paper, we propose a threat model and the request for requirements for developing NMTD techniques. For this purpose, we first examine the challenging problems in the NMTD mechanisms that were proposed for the legacy TCP/IP network. Secondly, we present a threat model in terms of attacker's intelligence, the intended information scope, and the attacker's location. Lastly, we provide seven basic requirements to develop an NMTD mechanism for the legacy TCP/IP network: 1) end-host address mutation, 2) post tracking, 3) address mutation unit, 4) service transparency, 5) name and address access, 6) adaptive defense, and 7) controller operation. We believe that this paper gives some insight into how to design and implement a new NMTD mechanism that would be deployable in real network.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권11호
• /
• pp.5673-5691
• /
• 2019

Information centric mobile network can significantly improve the data retrieving efficiency by caching contents at mobile edge. However, the cache pollution attack can affect the data obtaining process severely by requiring unpopular contents deliberately. To tackle the problem, we design an algorithm of mitigating cache pollution attacks in information centric mobile network. Particularly, the content popularity distribution statistic is proposed to detect abnormal behavior. Then a probabilistic caching strategy based on abnormal behavior is applied to dynamically maintain the steady-state distribution for content visiting probability and achieve the purpose of defense. The experimental results show that the proposed scheme can achieve higher request hit ratio and smaller latency for false locality content pollution attack than the CacheShield approach and the baseline approach where no mitigation approach is applied.

• International Journal of Computer Science & Network Security
• /
• 제24권4호
• /
• pp.179-191
• /
• 2024

With the advancement of modern technology, cyber-attacks are always rising. Specialized defense systems are needed to protect organizations against these threats. Malicious behavior in the network is discovered using security tools like intrusion detection systems (IDS), firewall, antimalware systems, security information and event management (SIEM). It aids in defending businesses from attacks. Delivering advance threat feeds for precise attack detection in intrusion detection systems is the role of cyber-threat intelligence (CTI) in the study is being presented. In this proposed work CTI feeds are utilized in the detection of assaults accurately in intrusion detection system. The ultimate objective is to identify the attacker behind the attack. Several data sets had been analyzed for attack detection. With the proposed study the ability to identify network attacks has improved by using machine learning algorithms. The proposed model provides 98% accuracy, 97% precision, and 96% recall respectively.

• 정보처리학회논문지C
• /
• 제14C권1호
• /
• pp.45-54
• /
• 2007

가장 위협적인 공격의 한 형태인 DDoS(Distributed Denial of Service) 공격은 다수의 공격 에이전트가 한꺼번에 많은 공격 트래픽을 특정 네트워크 또는 중요한 노드를 공격하는 특성을 갖고 있어 이로 인한 피해 지역 및 정도가 크다는 문제점이 있다. 이에 대한 기존의 많은 연구들은 탐지, 필터링, 추적 등에 집중되어 있고, 특히 피해 네트워크가 계층적인 구조를 갖고 있는 경우 특정 노드의 마비로 인해 하위 노드들의 정상 트래픽 전송이 어려워질 뿐 아니라, 탐지에 대한 다른 노드에 공지 및 추적을 위한 제어 트래픽 전송 또한 어려워 질 수 있다. 이에 본 논문에서는 계층적인 네트워크에서 이에 맞는 계층적인 오버레이를 구성하여, 공격 탐지 시 공지 및 추적을 위한 제어 트래픽을 오버레이를 이용해 전달하며, 공격 에이전트를 완전히 제거하기 전까지 정상적인 트래픽을 우회할 수 있는 DDoS 공격 감내 네트워크 구조를 제안한다. 또한 제안된 방법에서 오버레이 구성에 따른 오버헤드 분석과 공격 탐지 시 빠른 공격 차단 전달의 가능성과 신속성 및 정상 트래픽의 전송의 정도를 시뮬레이션을 통해 분석한다.

• 한국컴퓨터정보학회논문지
• /
• 제18권8호
• /
• pp.95-104
• /
• 2013

2009년 77DDoS 대란 이후 DDoS 공격은 사회적 위협으로 발전하고 있다. 이러한 위협에 대응하기 위해 다양한 DDoS방어기법이 연구되고 있으나, DDoS 공격기법 또한 더욱 정교해지고 있다. DDoS 공격의 형태는 과거 네트워크계층의 대용량 트래픽 공격에서 최근에는 애플리케이션 계층의 소량의 정교한 형태(Slow DDoS Attack)로 변하고 있으며 공격을 위한 공격에이전트 또한 더욱 지능화 되고 정상 PC와 구분이 모호하여 차단이 더욱 어렵게 되고 있다. 정상PC와 지능화된 공격에이전트 구분을 위해 최근 사용되는 사용자인증시스템(CAPTCHA)의 경우 인증과정에서 사용자의 개입이 필요하며, 특히 NAT 환경에서 IP 기반 차단 방법은 정상사용자의 트래픽까지 동시 차단될 수 있다. 본 논문에서는 HTTP 프로토콜에서 사용하는 쿠키를 활용한 일회성 세션 키 기반 인증방법을 적용하여 공격 에이전트와 정상 PC를 구분, HTTP DDoS 공격을 효과적으로 차단하기 위한 방어기법을 제안한다.

• KEPCO Journal on Electric Power and Energy
• /
• 제1권1호
• /
• pp.55-59
• /
• 2015

Denial of Service (DoS) attack is to interfere the normal user from using the information technology services. With a rapid technology improvements in computer and internet environment, small sized DoS attacks targeted to server or network infrastructure have been disabled. Thus, Distributed Denial of Service (DDoS) attacks that utilizes from tens to several thousands of distributed computers as zombie PC appear to have as one of the most challenging threat. In this paper, we categorize the DDoS attacks and classify existing countermeasures based on where and when they prevent, detect, and respond to the DDoS attacks. Then we propose a comprehensive defense mechanism against DDoS attacks in Control System to detect attacks efficiently.