Distributed Denial of Service (DDoS) detection by traffic pattern analysis

In this paper, we propose a behavior-based detection that can discriminate Distributed Denial of Service (DDoS) attack traffic from legitimated traffic regardless to various types of the attack packets and methods. Current DDoS attacks are carried out by attack tools, worms and botnets using different packet-transmission rates and packet forms to beat defense systems. These various attack strategies lead to defense systems requiring various detection methods in order to identify the attacks. Moreover, DDoS attacks can craft the traffics like flash crowd events and fly under the radar through the victim. We notice that DDoS attacks have features of repeatable patterns which are different from legitimate flash crowd traffics. In this paper, we propose a comparable detection methods based on the Pearson’s correlation coefficient. Our methods can extract the repeatable features from the packet arrivals in the DDoS traffics but not in flash crowd traffics. The extensive simulations were tested for the optimization of the detection methods. We then performed experiments with several datasets and our results affirm that the proposed methods can differentiate DDoS attacks from legitimate traffics.     

Detection of DDoS attacks using optimized traffic matrix

Distributed Denial of Service (DDoS) attacks have been increasing with the growth of computer and network infrastructures in Ubiquitous computing. DDoS attacks generating mass traffic deplete network bandwidth and/or system resources. It is therefore significant to detect DDoS attacks in their early stage. Our previous approach used a traffic matrix to detect DDoS attacks quickly and accurately. However, it could not find out to tune up parameters of the traffic matrix including (i) size of traffic matrix, (ii) time based window size, and (iii) a threshold value of variance from packets information with respect to various monitored environments and DDoS attacks. Moreover, the time based window size led to computational overheads when DDoS attacks did not occur. To cope with it, we propose an enhanced DDoS attacks detection approach by optimizing the parameters of the traffic matrix using a Genetic Algorithm (GA) to maximize the detection rates. Furthermore, we improve the traffic matrix building operation by (i) reforming the hash function to decrease hash collisions and (ii) replacing the time based window size with a packet based window size to reduce the computational overheads. We perform experiments with DARPA 2000 LLDOS 1.0, LBL-PKT-4 of Lawrence Berkeley Laboratory and generated attack datasets. The experimental results show the feasibility of our approach in terms of detection accuracy and speed.     

云环境下一种基于信任的加密流量DDoS发现方法

针对云环境下分布式拒绝服务(distributed denial-of-service,DDoS)攻击加密攻击流量隐蔽性更强、更容易发起、规模更大的问题,提出了一种云环境下基于信任的加密流量DDoS发现方法TruCTCloud.该方法在现有基于机器学习的DDoS攻击检测中引入信任的思想,结合云服务自身的安全认证,融入基...     

基于地址相关度的分布式拒绝服务攻击检测方法

分布式拒绝服务(DDoS)攻击检测是网络安全领域的研究热点.对DDoS攻击的研究进展及其特点进行了详细分析,针对DDoS攻击流的流量突发性、流非对称性、源IP地址分布性和目标IP地址集中性等本质特征提出了网络流的地址相关度(ACV)的概念.为了充分利用ACV,提高方法的检测质量,提出了基于ACV的DDoS攻击检测方法,通过自回归模型的参数拟合将ACV时间序列变换为多维空间内的AR模型参数向量序列来描述网络流状态特征,采用支持向量机分类器对当前网络流状态进行分类以识别DDoS攻击.实验结果表明,该检测方法能够有效地检测DDoS攻击,降低误报率.     

基于统计分析的DDoS攻击检测

分析了分布式拒绝服务（Distributed Denial of Service,DDoS）攻击原理及其攻击特征,从提高检测响应时间和减少计算复杂性的角度提出了一种新的DDoS攻击检测方法。该方法基于DDoS攻击的固有特性,从IP连接数据的统计分析中寻找能够描述系统正常行为的分布规律,建立基于统计分析的DDoS攻击检测模型。实验结果表明,该方法能快速有效地实现对DDoS攻击的检测,并对其他网络安全检测具有指导作用。     

A router-based technique to mitigate reduction of quality (RoQ) attacks

We propose a router-based technique to mitigate the stealthy reduction of quality (RoQ) attacks at the routers in the Internet. The RoQ attacks have been shown to impair the QoS sensitive VoIP and the TCP traffic in the Internet. It is difficult to detect these attacks because of their low average rates. We also show that our generalized approach can detect these attacks even if they employ the source IP address spoofing, the destination IP address spoofing, and undefined periodicity to evade several router-based detection systems. The detection system operates in two phases: in phase 1, the presence of the RoQ attack is detected from the readily available per flow information at the routers, and in phase 2, the attack filtering algorithm drops the RoQ attack packets. Assuming that the attacker uses the source IP address and the destination IP address spoofing, we propose to detect the sudden increase in the traffic load of all the expired flows within a short period. In a network without RoQ attacks, we show that the traffic load of all the expired flows is less than certain thresholds, which are derived from real Internet traffic analysis. We further propose a simple filtering solution to drop the attack packets. The filtering scheme treats the long-lived flows in the Internet preferentially, and drops the attack traffic by monitoring the queue length if the queue length exceeds a threshold percent of the queue limit. Our results show that we can successfully detect and mitigate RoQ attacks even with the source and destination IP addresses spoofed. The detection system is implemented in the ns2 simulator. In the simulations, we use the flowid field available in ns2 to implement per-flow logic, which is a combination of the source IP address, the destination IP address, the source port, and the destination port. We also discuss the real implementation of the proposed detection system.     

You can run, but you can't hide: an effective statistical methodology to trace back DDoS attackers

There is currently an urgent need for effective solutions against distributed denial-of-service (DDoS) attacks directed at many well-known Web sites. Because of increased sophistication and severity of these attacks, the system administrator of a victim site needs to quickly and accurately identify the probable attackers and eliminate the attack traffic. Our work is based on a probabilistic marking algorithm in which an attack graph can be constructed by a victim site. We extend the basic concept such that one can quickly and efficiently deduce the intensity of the "local traffic" generated at each router in the attack graph based on the volume of received marked packets at the victim site. Given the intensities of these local traffic rates, we can rank the local traffic and identify the network domains generating most of the attack traffic. We present our trace back and attacker identification algorithms. We also provide a theoretical framework to determine the minimum stable time t/sub min/, which is the minimum time needed to accurately determine the locations of attackers and local traffic rates of participating routers in the attack graph. Extensive experiments are carried out to illustrate that one can accurately determine the minimum stable time t/sub min/ and, at the same time, determine the location of attackers under various threshold parameters, network diameters, attack traffic distributions, on/off patterns, and network traffic conditions.     

Building an Identity Management Infrastructure for Today…and Tomorrow

ABSTRACT

The basis of denial of service (DoS)/distributed DoS (DDoS) attacks lies in overwhelming a victim's computer resources by flooding them with enormous traffic. This is done by compromising multiple systems that send a high volume of traffic. The traffic is often formulated in such a way that it consumes finite resources at abnormal rates either at victim or network level. In addition, spoofing of source addresses makes it difficult to combat such attacks. This paper adopts a twofold collaborative mechanism, wherein the intermediate routers are engaged in markings and the victim uses these markings for detecting and filtering the flooding attacks. The markings are used to distinguish the legitimate network traffic from the attack so as to enable the routers near the victim to filter the attack packets. The marked packets are also helpful to backtrack the true origin of the spoofed traffic, thus dropping them at the source rather than allowing them to traverse the network. To further aid in the detection of spoofed traffic, Time to Live (TTL) in the IP header is used. The mappings between the IP addresses and the markings along with the TTLs are used to find the spurious traffic. We provide numerical and simulated experimental results to show the effectiveness of the proposed system in distinguishing the legitimate traffic from the spoofed. We also give a statistical report showing the performance of our system.     

Collaborative detection and filtering of shrew DDoS attacks using spectral analysis

This paper presents a new spectral template-matching approach to countering shrew distributed denial-of-service (DDoS) attacks. These attacks are stealthy, periodic, pulsing, and low-rate in attack volume, very different from the flooding type of attacks. They are launched with high narrow spikes in very low frequency, periodically. Thus, shrew attacks may endanger the victim systems for a long time without being detected. In other words, such attacks may reduce the quality of services unnoticeably. Our defense method calls for collaborative detection and filtering (CDF) of shrew DDoS attacks. We detect shrew attack flows hidden in legitimate TCP/UDP streams by spectral analysis against pre-stored template of average attack spectral characteristics. This novel scheme is suitable for either software or hardware implementation.The CDF scheme is implemented with the NS-2 network simulator using real-life Internet background traffic mixed with attack datasets used by established research groups. Our simulated results show high detection accuracy by merging alerts from cooperative routers. Both theoretical modeling and simulation experimental results are reported here. The experiments achieved up to 95% successful detection of network anomalies along with a low 10% false positive alarms. The scheme cuts off malicious flows containing shrew attacks using a newly developed packet-filtering scheme. Our filtering scheme retained 99% of legitimate TCP flows, compared with only 20% TCP flows retained by using the Drop Tail algorithm. The paper also considers DSP, FPGA, and network processor implementation issues and discusses limitations and further research challenges.     

基于改进AAR模型的DIDoS攻击早期检测方法

分布式增速拒绝服务(DIDoS)攻击采用逐步提升发包速率的方式来造成受害者资源的慢消耗,较之传统的分布式拒绝服务(DDoS)攻击更具隐蔽性,如何尽可能早地将其捕获是一个亟待研究的问题。本文针对DIDoS攻击的特点,提出了一种基于改进AAR模型的DIDoS攻击早期检测方法。为此,首先提出了一组基于条件熵的检测特征:流特征条件熵(TFCE),用以反映DIDoS攻击流速的增长变化;然后根据改进的AAR模型对TFCE值进行多步预测;最后采用经过训练的SVM分类器对预测值进行分类,以识别攻击企图。实验结果表明,在保证检测精度相当的前提下,该方法比部分现有方法能够更快检测到攻击。     

结合概率图模型与DNN的DDoS攻击检测方法

从传统网络到物联网,分布式拒绝服务攻击一直是网络安全的隐患。为提高分布式拒绝服务攻击的检测率,提出基于概率图模型与深度神经网络的DDoS攻击检测方案。该检测方案由数据预处理阶段和攻击检测阶段组成,在数据预处理阶段,研究了正常数据包与攻击包的区别,分别从TCP、UDP以及IP数据包包头信息提取出较高维的统计特征,根据随机森林计算的特征重要性因子,保留了前22个特征用于流量检测。22个统计特征通过概率图模型的隐马尔科夫算法进行聚类,然后将聚类结果通过检测阶段的深度神经网络对网络数据进行进一步的检测。在CICDoS数据集上进行验证性实验,结果表明,该检测方法的准确率最高可达99.35%,最低检测误报率和漏警率分别可达0.51%和0.12%。     

A Divide-and-Conquer Strategy for Thwarting Distributed Denial-of-Service Attacks

Attack mitigation schemes actively throttle attack traffic generated in distributed denial-of-service (DDoS) attacks. This paper presents attack diagnosis (AD), a novel attack mitigation scheme that adopts a divide-and-conquer strategy. AD combines the concepts of pushback and packet marking, and its architecture is in line with the ideal DDoS attack countermeasure paradigm - attack detection is performed near the victim host and packet filtering is executed close to the attack sources. AD is a reactive defense mechanism that is activated by a victim host after an attack is detected. By instructing its upstream routers to mark packets deterministically, the victim can trace back one attack source and command an AD-enabled router close to the source to filter the attack packets. This process isolates one attacker and throttles it, which is repeated until the attack is mitigated. We also propose an extension to AD called parallel attack diagnosis (PAD) that is capable of throttling traffic coming from a large number of attackers simultaneously. AD and PAD are analyzed and evaluated using the Skitter Internet map, Lumeta's Internet map, and the 6-degree complete tree topology model. Both schemes are shown to be robust against IP spoofing and to incur low false positive ratios     

基于流量优化的包标记IP追踪策略研究

传统的攻击源追踪方案在面对大规模DDoS攻击时,重构路径的收敛速度往往过慢.文中提出一种根据DDoS流量分布优化的随机包标记策略OMS(Optimized Marking Scheme),该策略通过在IP报头中插入控制信息,使标记包采样概率在攻击路径上随终点的距离递增,从而更远处的标记包能够以更高的概率到达终点.仿真试验的结果表明,OMS收敛速度较以往的方案有了明显的提高.     

基于线性预测的DDoS攻击检测方法

分布式拒绝服务攻击的原理简单、危害严重,如TCP淹没攻击。该文介绍一种快速、有效的方法来检测TCP SYN flooding攻击,通过线性预测分析来预防、拒绝服务攻击(DoS)。该检测机制采用TCP在响应超时情况下的指数回退算法性质,计算受攻击网络中的收到的SYN和发出的SYN＋ACK数据包数量之差进行数学建模,可以在很短的延时内检测SYN Flooding攻击。该算法可以方便地运用在叶节点路由器和防火墙中。     

DDoS攻击实时检测防御系统的硬件实现

分布式拒绝服务攻击是因特网安全的头号威胁。针对DDoS攻击,本文介绍了一种基于MPC860和FPGA的实时检测防御系统的体系结构与实现原理,探讨了基于非参数累积和(CUSUM)算法检测新IP地址到达速率变化的DDoS攻击检测方法。实验结果表明该系统不仅实时检测准确性高、在线检测速度快、防御效果好,而且不损失网络信息吞吐量,保证了合法用户的正常访问。     

网络DDoS攻击流的小波分析与检测

将小波分析中的小波变换模极大方法用于检测分布式拒绝服务攻击引起的突发流量。在探讨如何运用小波模极大对突发流量进行判定的基础上,设计了一个检测突发攻击流量的方法,并对实际采集到的网络流量和仿真攻击流量的混合流作了计算机模拟验证。结果表明,当攻击流的突变幅度为正常流量的2倍￣3倍时,检测漏判率不超过5%;当攻击流的突变幅度提升为正常流量均值的3倍￣5倍时,检测漏判率不超过1%。攻击越强,检测漏判率越小。     

基于 HMM的分布式拒绝服务攻击检测方法

在分布式拒绝服务（DDoS）攻击时，网络中数据包的统计特征会显示出异常．检测这种异常是一项重要的任务．一些检测方法基于数据包速率的假设，然而这种假设在一些情况下是不合理的．另一些方法基于IP地址和数据报长度的统计特征，但这些方法在IP地址欺骗攻击时检测率急剧下降．提出了一种基于隐马尔可夫模型（HMM）的DDoS异常检测方法．该方法集成了4种不同的检测模型以对付不同类型的攻击．通过从数据包中提取TCP标志位，UDP端口和ICMP类型及代码等属性信息建立相应的TCP，UDP和ICMP的隐马尔可夫模型，用于描述正常情况下网络数据包序列的统计特征．然后用它来检测网络数据包序列，判断是否有DDoS攻击．实验结果显示该方法与其他同类方法相比通用性更好、检测率更高．     

基于多核学习的自适应DDoS攻击检测方法

分布式拒绝服务DDoS攻击是互联网安全的主要威胁之一。当前大多数检测方法采用单一特征,在大数据环境下不能有效地检测DDoS早期攻击。提出了一种基于多核学习的特征自适应DDoS攻击检测方法FADADM,根据DDoS攻击流量的突发性、地址的分布性以及通信双方的交互性定义了5个特征。基于集成学习框架,分别提出采用增大同类方差与异类均值差的比值IS/M和减少同类方差与异类均值差的比值RS/M的方式自适应地调整各特征值的权重,基于简单多核学习SimpleMKL模型训练出IS/M-SimpleMKL和RS/M-SimpleMKL 2种具有不同特性的多核学习模型,以识别DDoS早期攻击。实验结果表明,本文方法能够快速、准确地检测DDoS早期攻击。     

基于流量行为特征的DoS&DDoS攻击检测与异常流识别

针对现有方法仅分析粗粒度的网络流量特征参数,无法在保证检测实时性的前提下识别出拒绝服务(DoS)和分布式拒绝服务(DDoS)的攻击流这一问题,提出一种骨干网络DoS&DDoS攻击检测与异常流识别方法。首先,通过粗粒度的流量行为特征参数确定流量异常行为发生的时间点;然后,在每个流量异常行为发生的时间点对细粒度的流量行为特征参数进行分析,以找出异常行为对应的目的IP地址;最后,提取出与异常行为相关的流量进行综合分析,以判断异常行为是否为DoS攻击或者DDoS攻击。仿真实验的结果表明,基于流量行为特征的DoS&DDoS攻击检测与异常流识别方法能有效检测出骨干网络中的DoS攻击和DDoS攻击,并且在保证检测实时性的同时,准确地识别出与攻击相关的网络流量     

An autonomous defense against SYN flooding attacks: Detect and throttle attacks at the victim side independently

Distributed denial of service (DDoS) attacks seriously threaten Internet services yet there is currently no defence against such attacks that provides both early detection, allowing time for counteraction, and an accurate response. Traditional detection methods rely on passively sniffing an attacking signature and are inaccurate in the early stages of an attack. Current counteractions such as traffic filter or rate-limit methods do not accurately distinguish between legitimate and illegitimate traffic and are difficult to deploy. This work seeks to provide a method that detects SYN flooding attacks in a timely fashion and that responds accurately and independently on the victim side. We use the knowledge of network traffic delay distribution and apply an active probing technique (DARB) to identify half-open connections that, suspiciously, may not arise from normal network congestion. This method is suitable for large network areas and is capable of handling bursts of traffic flowing into a victim server. Accurate filtering is ensured by a counteraction method using IP address and time-to-live(TTL) fields. Simulation results show that our active detection method can detect SYN flooding attacks accurately and promptly and that the proposed rate-limit counteraction scheme can efficiently minimize the damage caused by DDoS attacks and guarantee constant services to legitimate users.