Optimization of Stealthwatch Network Security System for the Detection and Mitigation of Distributed Denial of Service (DDoS) Attack: Application to Smart Grid System

The Smart Grid is an enhancement of the traditional grid system and employs new technologies and sophisticated communication techniques for electrical power transmission and distribution. The Smart Grid’s communication network shares information about status of its several integrated IEDs (Intelligent Electronic Devices). However, the IEDs connected throughout the Smart Grid, open opportunities for attackers to interfere with the communications and utilities resources or take clients’ private data. This development has introduced new cyber-security challenges for the Smart Grid and is a very concerning issue because of emerging cyber-threats and security incidents that have occurred recently all over the world. The purpose of this research is to detect and mitigate Distributed Denial of Service [DDoS] with application to the Electrical Smart Grid System by deploying an optimized Stealthwatch Secure Network analytics tool. In this paper, the DDoS attack in the Smart Grid communication networks was modeled using Stealthwatch tool. The simulated network consisted of Secure Network Analytic tools virtual machines (VMs), electrical Grid network communication topology, attackers and Target VMs. Finally, the experiments and simulations were performed, and the research results showed that Stealthwatch analytic tool is very effective in detecting and mitigating DDoS attacks in the Smart Grid System without causing any blackout or shutdown of any internal systems as compared to other tools such as GNS3, NeSSi2, NISST Framework, OMNeT++, INET Framework, ReaSE, NS2, NS3, M5 Simulator, OPNET, PLC & TIA Portal management Software which do not have the capability to do so. Also, using Stealthwatch tool to create a security baseline for Smart Grid environment, contributes to risk mitigation and sound security hygiene.

The History, Trend, Types, and Mitigation of Distributed Denial of Service Attacks

Over time, the world has transformed digitally and there is total dependence on the internet. Many more gadgets are continuously interconnected in the internet ecosystem. This fact has made the Internet a global information source for every being. Despite all this, attacker knowledge by cybercriminals has advanced and resulted in different attack methodologies on the internet and its data stores. This paper will discuss the origin and significance of Denial of Service (DoS) and Distributed Denial of Service (DDoS). These kinds of attacks remain the most effective methods used by the bad guys to cause substantial damage in terms of operational, reputational, and financial damage to organizations globally. These kinds of attacks have hindered network performance and availability. The victim’s network is flooded with massive illegal traffic hence, denying genuine traffic from passing through for authorized users. The paper will explore detection mechanisms, and mitigation techniques for this network threat.

Formalized Description of Distributed Denial of Service Attack

The distributed denial of service (DDoS) attack is one of the dangers in intrusion modes. It's difficult to defense and can cause serious damage to the system. Based on a careful study of the attack principles and characteristics, an object-oriented formalized description is presented, which contains a three-level framework and offers full specifications of all kinds of DDoS modes and their features and the relations between one another. Its greatest merit lies in that it contributes to analyzing, checking and judging DDoS. Now this formalized description has been used in a special IDS and it works very effectively.(

基于CNN-BiLSTM的ICMPv6 DDoS攻击检测方法

针对ICMPv6网络中DDoS攻击检测问题,提出一种基于CNN-BiLSTM网络的检测算法。通过将带有注意力机制、DropConnect和Dropout混合使用加入到CNN-BiLSTM算法中,防止在训练过程中产生的过拟合问题,同时更准确地提取数据的特性数据。通过实验表明:提出的算法在多次实验中的检测准确率、误报率与漏报率平均值分别为92.84%、4.49%和10.54%,检测算法泛化性较强,性能由于其他算法,能够有效处理ICMPv6 DDoS攻击检测问题。

面向边缘计算的TCA1C DDoS检测模型

边缘计算弥补了传统云计算数据传输开销大的不足,但边缘网络中存储和计算资源受限的特殊性限制了其部署复杂安全算法的能力,更易受到分布式拒绝服务(DDoS)攻击。针对目前边缘网络中DDoS攻击检测方法性能不高、未对卸载任务分类处理、对多属性的流量处理能力弱的问题,提出一种基于任务分类的Attention-1D-CNN DDoS检测模型TCA1C,对通信链路中的流量按不同的卸载任务进行分类,使单个任务受到攻击时不会影响整个链路中计算任务卸载的安全性,再对同一任务下的流量提取属性值并进行归一化处理。处理后的数据输入到Attention-1D-CNN,通道Attention和空间Attention学习数据特征对DDoS检测的贡献度,利用筛选函数剔除低于特征阈值的冗余信息,降低模型学习过程的复杂度,使模型快速收敛。仿真结果表明:TCA1C模型在缩短DDoS检测所用时间的情况下,检测准确率高达99.73%,检测性能优于DT、ELM、LSTM和CNN;当多个卸载任务在面临特定攻击概率时,卸载任务分类能有效降低不同任务的相互影响,使终端设备的计算任务在卸载过程中保持较高的安全性。

SDN中基于统计与集成自编码器的DDoS攻击检测模型

软件定义网络(Software-defined Networking,SDN)是一种提供细颗粒集中网络管理服务的新型网络体系结构,主要有控制与转发分离、集中控制和开放接口基本特征。SDN由于控制层的集中管理逻辑,控制器被攻击者作为理想的分布式拒绝服务攻击(Distributed Denial-of-Service,DDoS)目标。然而,传统的基于统计的DDoS攻击检测算法常存在误报率高、阈值固定等问题;基于机器学习模型的检测算法常存在计算资源消耗大、泛化性差等问题。为此,文中提出了一种基于统计特征与集成自编码器的DDoS攻击双层检测模型。基于统计的方法提取Rényi熵特征,设置动态阈值判断可疑流量;基于集成自编码器算法对可疑流量进行更精确的DDoS攻击判断。双层检测模型不仅提升了检测效果,解决了误报率高的问题,同时还有效地缩短了检测时间,从而减少了计算资源的消耗。实验结果表明,该模型在不同网络环境下都有较高的准确率,不同数据集检测的F1值最低都达到了98.5%以上,表现出了很强的泛化性。

Game-theoretical Model for Dynamic Defense Resource Allocation in Cyber-physical Power Systems Under Distributed Denial of Service Attacks

Electric power grids are evolving into complex cyber-physical power systems(CPPSs)that integrate advanced information and communication technologies(ICTs)but face increasing cyberspace threats and attacks.This study considers CPPS cyberspace security under distributed denial of service(DDoS)attacks and proposes a nonzero-sum game-theoretical model with incomplete information for appropriate allocation of defense resources based on the availability of limited resources.Task time delay is applied to quantify the expected utility as CPPSs have high time requirements and incur massive damage DDoS attacks.Different resource allocation strategies are adopted by attackers and defenders under the three cases of attack-free,failed attack,and successful attack,which lead to a corresponding consumption of resources.A multidimensional node value analysis is designed to introduce physical and cybersecurity indices.Simulation experiments and numerical results demonstrate the effectiveness of the proposed model for the appropriate allocation of defense resources in CPPSs under limited resource availability.

融合稀疏注意力机制在DDoS攻击检测中的应用

针对现有的DDoS(distributed denial of service)攻击检测模型面临大量数据时,呈现出检测效率低的问题。为适应当前网络环境,通过研究DDoS攻击检测模型、提取流量特征、计算攻击密度,提出一种基于融合稀疏注意力机制的DDoS攻击检测模型GVBNet(global variable block net),使用攻击密度自适应计算稀疏注意力。利用信息熵以及信息增益分析提取攻击流量的连续字节作为特征向量,通过构建基于GVBNet的网络模型在两种数据集上进行训练。实验结果表明,该方法具有良好的识别效果、检测速度以及抗干扰能力,在不同的环境下具有应用价值。

SDN中基于信息熵与机器学习的DDoS攻击检测模型构建

软件定义网络(Software-Defined Network,SDN)的集中控制特征使得网络管理更加灵活高效,但同时也成为网络攻击的主要对象,其中分布式拒绝服务攻击DDoS是SDN面临的主要威胁之一。结合统计学习和机器学习这2种SDN中常用的检测方法,文章分析了基于信息熵与机器学习算法的DDoS攻击检测模型,并利用信息熵的阈值判断检测出疑似异常流量,再用决策树算法构建的检测模型检测出DDoS攻击。分类检测模型构建了6个特征属性,并通过计算信息增益值筛选出最优特征子集。通过与其他分类算法模型的比较,该模型提高了检测准确性,减少了检测时间。

A Novel Framework for DDoS Attacks Detection Using Hybrid LSTM Techniques

The recent development of cloud computing offers various services on demand for organization and individual users,such as storage,shared computing space,networking,etc.Although Cloud Computing provides various advantages for users,it remains vulnerable to many types of attacks that attract cyber criminals.Distributed Denial of Service(DDoS)is the most common type of attack on cloud computing.Consequently,Cloud computing professionals and security experts have focused on the growth of preventive processes towards DDoS attacks.Since DDoS attacks have become increasingly widespread,it becomes difficult for some DDoS attack methods based on individual network flow features to distinguish various types of DDoS attacks.Further,the monitoring pattern of traffic changes and accurate detection of DDoS attacks are most important and urgent.In this research work,DDoS attack detection methods based on deep belief network feature extraction and Hybrid Long Short-Term Memory(LSTM)model have been proposed with NSL-KDD dataset.In Hybrid LSTM method,the Particle Swarm Optimization(PSO)technique,which is combined to optimize the weights of the LSTM neural network,reduces the prediction error.This deep belief network method is used to extract the features of IP packets,and it identifies DDoS attacks based on PSO-LSTM model.Moreover,it accurately predicts normal network traffic and detects anomalies resulting from DDoS attacks.The proposed PSO-LSTM architecture outperforms the classification techniques including standard Support Vector Machine(SVM)and LSTM in terms of attack detection performance along with the results of the measurement of accuracy,recall,f-measure,precision.

基于深度森林的多类型DDoS攻击检测方法

分布式拒绝服务攻击(DDoS)是网络安全的主要威胁之一。近年来,基于多种不同DDoS攻击方式的混合攻击数量大幅增长,如何在保证精度的前提下同时检测多种类型的DDoS攻击成为亟待解决的问题。为此,提出一种基于深度森林的多类型DDoS攻击检测方法。该方法首先使用基于平均不纯度的特征选择算法对多类型异常流量数据集进行特征排序与特征筛选;然后使用多粒度扫描对DDoS训练集进行特征提取,并使用级联森林分层训练模型,最终生成可用于DDoS恶意流量检测与分类的深度森林模型。实验结果表明,与6种主流树类集成学习模型相比,基于改进深度森林的DDoS攻击检测方法训练得到的分类器准确率最低提升了0.8%,召回率最低提升了0.9%;与改进前相比,改进后模型准确率提升了1.3%,加权召回率提高了1.3%,训练时间减少了29.7%。模型整体性能有明显提升。

计算机网络中的DDoS攻击检测与防御机制研究

随着互联网的迅猛发展,计算机网络已成为社会和经济中不可或缺的一部分。然而,人们对网络的过度依赖为恶意攻击者提供了机会,其中分布式拒绝服务(DDoS)攻击是一种常见且严重的威胁。文章深入研究了DDoS攻击的检测与防御机制,旨在为网络提供更加全面、有效的保护。

Detecting and Mitigating DDOS Attacks in SDNs Using Deep Neural Network

Distributed denial of service(DDoS)attack is the most common attack that obstructs a network and makes it unavailable for a legitimate user.We proposed a deep neural network(DNN)model for the detection of DDoS attacks in the Software-Defined Networking(SDN)paradigm.SDN centralizes the control plane and separates it from the data plane.It simplifies a network and eliminates vendor specification of a device.Because of this open nature and centralized control,SDN can easily become a victim of DDoS attacks.We proposed a supervised Developed Deep Neural Network(DDNN)model that can classify the DDoS attack traffic and legitimate traffic.Our Developed Deep Neural Network(DDNN)model takes a large number of feature values as compared to previously proposed Machine Learning(ML)models.The proposed DNN model scans the data to find the correlated features and delivers high-quality results.The model enhances the security of SDN and has better accuracy as compared to previously proposed models.We choose the latest state-of-the-art dataset which consists of many novel attacks and overcomes all the shortcomings and limitations of the existing datasets.Our model results in a high accuracy rate of 99.76%with a low false-positive rate and 0.065%low loss rate.The accuracy increases to 99.80%as we increase the number of epochs to 100 rounds.Our proposed model classifies anomalous and normal traffic more accurately as compared to the previously proposed models.It can handle a huge amount of structured and unstructured data and can easily solve complex problems.

Iterative Dichotomiser Posteriori Method Based Service Attack Detection in Cloud Computing

Cloud computing(CC)is an advanced technology that provides access to predictive resources and data sharing.The cloud environment represents the right type regarding cloud usage model ownership,size,and rights to access.It introduces the scope and nature of cloud computing.In recent times,all processes are fed into the system for which consumer data and cache size are required.One of the most security issues in the cloud environment is Distributed Denial of Ser-vice(DDoS)attacks,responsible for cloud server overloading.This proposed sys-tem ID3(Iterative Dichotomiser 3)Maximum Multifactor Dimensionality Posteriori Method(ID3-MMDP)is used to overcome the drawback and a rela-tively simple way to execute and for the detection of(DDoS)attack.First,the pro-posed ID3-MMDP method calls for the resources of the cloud platform and then implements the attack detection technology based on information entropy to detect DDoS attacks.Since because the entropy value can show the discrete or aggregated characteristics of the current data set,it can be used for the detection of abnormal dataflow,User-uploaded data,ID3-MMDP system checks and read risk measurement and processing,bug ratingfile size changes,orfile name changes and changes in the format design of the data size entropy value.Unique properties can be used whenever the program approaches any data error to detect abnormal data services.Finally,the experiment also verifies the DDoS attack detection capability algorithm.

Denial of Service Due to Direct and Indirect ARP Storm Attacks in LAN Environment

ARP-based Distributed Denial of Service (DDoS) attacks due to ARP-storms can happen in local area networks where many computer systems are infected by worms such as Code Red or by DDoS agents. In ARP attack, the DDoS agents constantly send a barrage of ARP requests to the gateway, or to a victim computer within the same sub-network, and tie up the resource of attacked gateway or host. In this paper, we set to measure the impact of ARP-attack on resource exhaustion of computers in a local area network. Based on attack experiments, we measure the exhaustion of processing and memory resources of a victim computer and also other computers, which are located on the same network as the victim computer. Interestingly enough, it is observed that an ARP-attack not only exhausts resource of the victim computer but also significantly exhausts processing resource of other non-victim computers, which happen to be located on the same local area network as the victim computer.

基于机器学习的无线网络DDoS攻击检测方法

为提高分布式拒绝服务(Distributed Denial of Service,DDoS)攻击检出率,设计基于机器学习的无线网络DDoS攻击检测方法。首先,结合攻击时间序列构建无线网络DDoS攻击检测模型,利用深度学习设计无线网络DDoS攻击检测机制;其次,通过异常流量判断,对照相应的流表特征信息完成分类检测;最后,进行实验分析。实验结果表明,该方法的DDoS攻击检出率较低,优于对照组。

SDN环境中基于Bi-LSTM的DDoS攻击检测方案

针对5G物联网环境中海量接入设备带来的DDoS攻击威胁,同时考虑到软件定义网络SDN对5G物联网的适用性,提出了一种在SDN环境中利用长短期记忆LSTM网络检测DDoS攻击的方案,以提高对DDoS攻击检测的准确性。并采用分治算法思想,提出了一种轻量级分布式边缘计算架构OCM,在物联网中的空闲边缘节点部署基于Bi-LSTM的轻量级神经网络完成检测任务,在保证准确性的同时,增加了检测的灵活性。在ISCX2012数据集上评估了所提方案的有效性和可行性。实验结果表明,所提方案能够准确检测DDoS攻击并有效缓解DDoS攻击。

边缘计算环境下基于深度学习的DDos检测

边缘计算作为一种用于降低中心节点计算压力,更靠近终端设备和数据源头的新计算范式,满足了计算业务下沉的需求,也带来了安全问题;其中,对边缘计算安全威胁最大、造成过巨大经济损失和安全事故的当属分布式拒绝服务攻击(DDos);边缘计算环境下由于算力受限、存储空间有限等原因,传统的防御手段难以应用;因此,提出了一种适用于边缘计算环境下的基于深度学习的轻量级DDos检测框架;采用CIC-DDos-2019数据集来模拟边缘计算环境下的遭受DDos攻击的网络流量,针对数据集进行了适应性强的预处理技术和相似性标签融合,运用SMOTE算法解决了数据集类别不平衡问题,采用一维卷积技术和BiLSTM技术搭建了模型并进行了模型剪枝,构建了一个轻量级模型;结果表明,其针对DDos攻击类别的八分类实验准确率达到了96.8%,二分类实验准确率达到了99.8%。

V2G网络中基于联邦学习和CNN-BiLSTM的DDoS攻击检测

DDoS攻击是V2G网络的重要威胁之一,它可以在短时间内耗尽服务器的通信资源。此前方法以集中式模型为主,将数据从边缘设备传输到中央服务器进行训练可能会将数据暴露给各种攻击。研究了一种基于联邦学习的入侵检测系统,首先,考虑到V2G网络数据的高维性和数据间的时间依赖性,将采集的数据通过改进的特征选择算法进行降维,减少冗余特征,再将处理后的数据传入到融合了卷积神经网络和双向长短时记忆网络的混合模型中,捕获数据中的时间依赖性,并引入批标准化防止神经网络训练过程中出现梯度消失问题;其次,为了防止隐私泄露,结合联邦学习的固有特性,允许数据留在本地用于神经网络模型的训练;为了解决联邦学习通信造成网络负载压力过大的问题,设计了一种通过设置动态通信阈值筛选参与更新最优边缘设备的方案以减轻网络负载压力。实验结果表明,该方法的准确率可以高达99.95%,单轮通信时间减少了1.7 s。

融合特征选择的随机森林DDoS攻击检测

现有基于机器学习的分布式拒绝服务(DDoS)攻击检测方法在面对愈发复杂的网络流量、不断升维的数据结构时,检测难度和成本不断上升。针对这些问题,提出一种融合特征选择的随机森林DDoS攻击检测方法。该方法选用基于基尼系数的平均不纯度算法作为特征选择算法,对DDoS异常流量样本进行降维,以降低训练成本、提高训练精度;同时将特征选择算法嵌入随机森林的单个基学习器,将特征子集搜索范围由全部特征缩小到单个基学习器对应特征,在提高两种算法耦合性的同时提高了模型精度。实验结果表明,融合特征选择的随机森林DDoS攻击检测方法训练所得到的模型,在限制决策树棵数和训练样本数量的前提下,召回率相较于改进前提升21.8个百分点,F1-score值提升12.0个百分点,均优于传统的随机森林检测方案。