Dynamic Trust Model Based on Service Recommendation in Big Data

In big data of business service or transaction, it is impossible to provide entire information to both of services from cyber system, so some service providers made use of maliciously services to get more interests. Trust management is an effective solution to deal with these malicious actions. This paper gave a trust computing model based on service-recommendation in big data. This model takes into account difference of recommendation trust between familiar node and stranger node. Thus, to ensure accuracy of recommending trust computing, paper proposed a fine-granularity similarity computing method based on the similarity of service concept domain ontology. This model is more accurate in computing trust value of cyber service nodes and prevents better cheating and attacking of malicious service nodes. Experiment results illustrated our model is effective.     

Preventing the impact of selfish behavior under MANET using Neighbor Credit Value based AODV routing algorithm

Mobile Ad hoc Network (MANET) nodes exchange information using the multi-hop wireless communications without the need for any pre-existing infrastructure. Routing protocols of MANET are designed with an assumption that the nodes will cooperate in routing process. To achieve high throughput and reliable communication, the nodes are expected to cooperate with each other. Routing protocol plays a crucial role in an effective communication between nodes and operates on the assumption that the nodes are fully cooperative. Due to the open structure and limited battery-based energy in MANET, some nodes may not cooperate correctly or behave maliciously and such kind of misbehavior impacts the fairness, reliability and efficiency in MANET. Previous work addressed the ways to overcome these kinds of node misbehaviors and attacks. Most of the existing works need time to analyse the neighbor traffic and decide whether a neighbor is behaving maliciously or not. Further, the existing credit-based detection mechanisms may mark a genuine idle node as a malicious node. This work addresses a simple Neighbor Credit Value based AODV (NCV-AODV) routing algorithm for the detection of selfish behavior which avoids such false detection. The proposed idea is implemented in Ad hoc On Demand Distance Vector (AODV) routing protocol and an extensive analysis on the performance of the proposed detection mechanism against the selfish behavior of some MANET nodes are conducted.     

无线传感器网络中基于信誉的恶意节点检测

提出一种基于信誉的恶意节点检测方法——RMDMN,在分簇的网络结构基础上,对节点的行为属性(如丢包率、转发率、位置匹配信息等)和网络攻击进行建模,结合阈值比较法动态地更新节点信誉值并进行恶意节点判断.实验仿真显示,该方法具有一定的恶意节点检测能力.     

基于组合图像特征与分层节点搜索的回环检测方法

目的 文中通过提出一种新的回环解决方案,平衡回环检测系统的高准确率与高运行效率。方法 提出一种利用组合图像特征与分层节点搜索的新方法。首先,计算一种原始图像的下采样二值化全局特征和经过改进的ORB(oriented FAST and rotated BRIEF)局部特征,将其存入图像特征数据库。其次,引入一种分层节点搜索算法,在数据库中搜索与当前图像特征最相似的全局特征作为回环候选。最后,利用改进的ORB特征进行局部特征匹配,验证候选图像,确定回环检测结果。结果 使用该算法在3个不同的数据集上进行验证,测试中每次回环检测的平均处理时间仅需19 ms。结论 实验结果表明,该算法在运行效率、准确率、召回率等方面均达到了领域内的先进水平。     

An Intelligent Forwarding Strategy in SDN-Enabled Named-Data IoV

Internet of Vehicles (IoV), a rapidly growing technology for efficient vehicular communication and it is shifting the trend of traditional Vehicular Ad Hoc Networking (VANET) towards itself. The centralized management of IoV endorses its uniqueness and suitability for the Intelligent Transportation System (ITS) safety applications. Named Data Networking (NDN) is an emerging internet paradigm that fulfills most of the expectations of IoV. Limitations of the current IP internet architecture are the main motivation behind NDN. Software-Defined Networking (SDN) is another emerging networking paradigm of technology that is highly capable of efficient management of overall networks and transforming complex networking architectures into simple and manageable ones. The combination of the SDN controller, NDN, and IoV can be revolutionary in the overall performance of the network. Broadcast storm, due to the broadcasting nature of NDN, is a critical issue in NDN based on IoV. High speed and rapidly changing topology of vehicles in IoV creates disconnected link problem and add unnecessary transmission delay. In order to cop-up with the above-discussed problems, we proposed an efficient SDN-enabled forwarding mechanism in NDN-based IoV, which supports the mobility of the vehicle and explores the cellular network for the low latency control messages. In IoV environment, the concept of Edge Controller (EC) to maintain and manage the in-time and real-time vehicular topology is being introduced. A mathematical estimation model is also proposed in our work that assists the centralized EC and SDN to find not only the shortest and best path but also the most reliable and durable path. The naming scheme and in-network caching property of the NDN nodes reduce the delay. We used ndnSIM and NS-3 for the simulation experiment along with SUMO for the environment generation. The results of NDSDoV illustrate significant performance in terms of availability with limited routing overhead, minimized delay, retransmissions, and increased packet satisfaction ratio. Besides, we explored the properties of EC that contribute mainly in path failure minimization in the network.     

An Intrusion Detection Algorithm Based on Feature Graph

With the development of Information technology and the popularization of Internet, whenever and wherever possible, people can connect to the Internet optionally. Meanwhile, the security of network traffic is threatened by various of online malicious behaviors. The aim of an intrusion detection system (IDS) is to detect the network behaviors which are diverse and malicious. Since a conventional firewall cannot detect most of the malicious behaviors, such as malicious network traffic or computer abuse, some advanced learning methods are introduced and integrated with intrusion detection approaches in order to improve the performance of detection approaches. However, there are very few related studies focusing on both the effective detection for attacks and the representation for malicious behaviors with graph. In this paper, a novel intrusion detection approach IDBFG (Intrusion Detection Based on Feature Graph) is proposed which first filters normal connections with grid partitions, and then records the patterns of various attacks with a novel graph structure, and the behaviors in accordance with the patterns in graph are detected as intrusion behaviors. The experimental results on KDD-Cup 99 dataset show that IDBFG performs better than SVM (Supprot Vector Machines) and Decision Tree which are trained and tested in original feature space in terms of detection rates, false alarm rates and run time.     

Game-Oriented Security Strategy Against Hotspot Attacks for Internet of Vehicles

With the rapid development of mobile communication technology, the application of internet of vehicles (IoV) services, such as for information services, driving safety, and traffic efficiency, is growing constantly. For businesses with low transmission delay, high data processing capacity and large storage capacity, by deploying edge computing in the IoV, data processing, encryption and decision-making can be completed at the local end, thus providing real-time and highly reliable communication capability. The roadside unit (RSU), as an important part of edge computing in the IoV, fulfils an important data forwarding function and provides an interactive communication channel for vehicles and server providers. Additional computing resources can be configured to accommodate the computing requirements of users. In this study, a virtual traffic defense strategy based on a differential game is proposed to solve the security problem of user-sensitive information leakage when an RSU is attacked. An incentive mechanism encourages service vehicles within the hot range to send virtual traffic to another RSU. By attracting the attention of attackers, it covers the target RSU and protects the system from attack. Simulation results show that the scheme provides the optimal strategy for intelligent vehicles to transmit virtual data, and ensures the maximization of users’ interests.     

Malware Detection Using Decision Tree Based SVM Classifier for IoT

The development in Information and Communication Technology has led to the evolution of new computing and communication environment. Technological revolution with Internet of Things (IoTs) has developed various applications in almost all domains from health care, education to entertainment with sensors and smart devices. One of the subsets of IoT is Internet of Medical things (IoMT) which connects medical devices, hardware and software applications through internet. IoMT enables secure wireless communication over the Internet to allow efficient analysis of medical data. With these smart advancements and exploitation of smart IoT devices in health care technology there increases threat and malware attacks during transmission of highly confidential medical data. This work proposes a scheme by integrating machine learning approach and block chain technology to detect malware during data transmission in IoMT. The proposed Machine Learning based Block Chain Technology malware detection scheme (MLBCT-Mdetect) is implemented in three steps namely: feature extraction, Classification and blockchain. Feature extraction is performed by calculating the weight of each feature and reduces the features with less weight. Support Vector Machine classifier is employed in the second step to classify the malware and benign nodes. Furthermore, third step uses blockchain to store details of the selected features which eventually improves the detection of malware with significant improvement in speed and accuracy. ML-BCT-Mdetect achieves higher accuracy with low false positive rate and higher True positive rate.     

Dynamic trust management for secure communications in social internet of things (SIoT)

The world has faced three Information and Communication Technology (ICT) revolutions and the third ICT wave led to Internet of Things, the notion of anything, everything, anytime and everywhere. Out of the many visions of IoT, one revolutionary concept is to make IoT sociable i.e., incorporating social networking within Internet of Things. This revolution has led to the notion of Social Internet of Things (SIoT). Establishing a SIoT network or community is not so simple and requires integration of heterogeneous technology and communication solutions. This paper focuses on establishing a secure and reliable communication over nodes in SIoT by computing trust dynamically among neighboring nodes. Trust Management is an important area that has attracted numerous researchers over the past few years. The proposed DTrustInfer computes trust based on first hand observation, second hand observation, centrality and dependability factor of a node. Properties of trust such as honesty, cooperativeness, community interest and energy of a node are considered for computing trust. Also, this paper ensures secure communication among SIoT nodes through simple secret codes. Experimental results show that the proposed DTrustInfer outperforms the existing trust models significantly.     

Enhanced Energy Efficient with a Trust Aware in MANET for Real-Time Applications

Mobile ad hoc networks (MANETs) are subjected to attack detection for transmitting and creating new messages or existing message modifications. The attacker on another node evaluates the forging activity in the message directly or indirectly. Every node sends short packets in a MANET environment with its identifier, location on the map, and time through beacons. The attackers on the network broadcast the warning message using faked coordinates, providing the appearance of a network collision. Similarly, MANET degrades the channel utilization performance. Performance highly affects network performance through security algorithms. This paper developed a trust management technique called Enhanced Beacon Trust Management with Hybrid Optimization (EBTM-Hyopt) for efficient cluster head selection and malicious node detection. It tries to build trust among connected nodes and may improve security by requiring every participating node to develop and distribute genuine, accurate, and trustworthy material across the network. Specifically, optimized cluster head election is done periodically to reduce and balance the energy consumption to improve the lifetime network. The cluster head election optimization is based on hybridizing Particle Swarm Optimization (PSO) and Gravitational Search Optimization Algorithm (GSOA) concepts to enable and ensure reliable routing. Simulation results show that the proposed EBTM-HYOPT outperforms the state-of-the-art trust model in terms of 297.99 kbps of throughput, 46.34% of PDR, 13% of energy consumption, 165.6 kbps of packet loss, 67.49% of end-to-end delay, and 16.34% of packet length.     

E2-SCAN: an extended credit strategy-based energy-efficient security scheme for wireless ad hoc networks

Utilising the battery life and the limited bandwidth available in mobile ad hoc networks (MANETs) in the most efficient manner is an important issue, along with providing security at the network layer. The authors propose, design and describe E2-SCAN, an energy-efficient network layered security solution for MANETs, which protects both routing and packet forwarding functionalities in the context of the on demand distance vector protocol. E2-SCAN is an advanced approach that builds on and improves upon some of the state-of-the-art results available in the literature. The proposed E2-SCAN algorithm protects the routing and data forwarding operations through the same reactive approach, as is provided by the SCAN algorithm. It also enhances the security of the network by detecting and reacting to the malicious nodes. In E2-SCAN, the immediate one-hop neighbour nodes collaboratively monitor. E2-SCAN adopts a modified novel credit strategy to decrease its overhead as the time evolves. Through both analysis and simulation results, the authors demonstrate the effectiveness of E2-SCAN over SCAN in a hostile environment.     

Enhanced multi-attribute trust protocol for malicious node detection in wireless sensor networks

There has been a tremendous growth in the field of wireless sensor networks (WSNs) in recent years, which is reflected in various applications. As the use of WSN applications increases, providing security to WSNs becomes a leading issue. This is complex due to the unique features of WSNs. This paper proposes a trust-based intrusion detection that uses multi-attribute trust metrics to improve detection accuracy. It uses an enhanced distributive trust calculation algorithm that involves monitoring neighbouring nodes and trust calculation using the trust metrics message success rate (MSR), elapsed time at node (ETN), correctness (CS) and fairness (FS). In addition to the normal communication-based trust property MSR, this paper uses effective parameters like ETN, which focuses on data and address modification attacks in an effective manner, and two social-interaction-based parameters CS and FS, which address trust-related attacks effectively. Simulation results show that the proposed method has higher performance and provides more security in terms of detection accuracy and false alarm rate.     

Energy Theft Detection in Smart Grids with Genetic Algorithm-Based FeatureSelection

As big data, its technologies, and application continue to advance, the Smart Grid (SG) has become one of the most successful pervasive and fixed computing platforms that efficiently uses a data-driven approach and employs efficient information and communication technology (ICT) and cloud computing. As a result of the complicated architecture of cloud computing, the distinctive working of advanced metering infrastructures (AMI), and the use of sensitive data, it has become challenging to make the SG secure. Faults of the SG are categorized into two main categories, Technical Losses (TLs) and Non-Technical Losses (NTLs). Hardware failure, communication issues, ohmic losses, and energy burnout during transmission and propagation of energy are TLs. NTL’s are human-induced errors for malicious purposes such as attacking sensitive data and electricity theft, along with tampering with AMI for bill reduction by fraudulent customers. This research proposes a data-driven methodology based on principles of computational intelligence as well as big data analysis to identify fraudulent customers based on their load profile. In our proposed methodology, a hybrid Genetic Algorithm and Support Vector Machine (GA-SVM) model has been used to extract the relevant subset of feature data from a large and unsupervised public smart grid project dataset in London, UK, for theft detection. A subset of 26 out of 71 features is obtained with a classification accuracy of 96.6%, compared to studies conducted on small and limited datasets.     

一种基于降维密度聚类的船舶异常轨迹识别方法

目的 有效分析和探索海洋船舶时空轨迹行为模式,提高船舶轨迹聚类的效率与质量,更好地检测真实船舶的异常行为。方法 针对当前船舶轨迹数据研究中存在的对多维特征信息利用不足、检测效率不高、检测精度较差等问题,提出一种精确度高、能自主识别分析多维特征的船舶异常轨迹识别方法。首先利用随机森林分类器评估多维特征重要性,构建轨迹特征的最优组合;然后提出一种降维密度聚类方法,将T–分布随机邻域嵌入(T–SNE)和自适应密度聚类(DBSCAN)模型结合,通过构建特征选择层和无监督聚类层实现对数据元素非线性关系的高效提取以及对聚类参数的智能选择;最后根据聚类结果构建类簇特征向量,计算距离阈值判别轨迹相似度,实现轨迹异常检测模型的构建。结果 以UCI数据集为例,降维密度聚类方法对4、13、30、64维特征数据集的F1分数能达到0.9 048、0.9 534、0.8 218、0.6 627,多个聚类指标均优于DBSCAN、K–Means等常见聚类算法的。结论 研究结果表明,降维密度聚类方法能有效提取数据多维特征结构,实现聚类参数自适应,弥补密度聚类中参数难以确定的问题,有效实现对多种类型船舶轨迹异常的识别。     

Identification of Anomaly Scenes in Videos Using Graph Neural Networks

Generally, conventional methods for anomaly detection rely on clustering, proximity, or classification. With the massive growth in surveillance videos, outliers or anomalies find ingenious ways to obscure themselves in the network and make conventional techniques inefficient. This research explores the structure of Graph neural networks (GNNs) that generalize deep learning frameworks to graph-structured data. Every node in the graph structure is labeled and anomalies, represented by unlabeled nodes, are predicted by performing random walks on the node-based graph structures. Due to their strong learning abilities, GNNs gained popularity in various domains such as natural language processing, social network analytics and healthcare. Anomaly detection is a challenging task in computer vision but the proposed algorithm using GNNs efficiently performs the identification of anomalies. The Graph-based deep learning networks are designed to predict unknown objects and outliers. In our case, they detect unusual objects in the form of malicious nodes. The edges between nodes represent a relationship of nodes among each other. In case of anomaly, such as the bike rider in Pedestrians data, the rider node has a negative value for the edge and it is identified as an anomaly. The encoding and decoding layers are crucial for determining how statistical measurements affect anomaly identification and for correcting the graph path to the best possible outcome. Results show that the proposed framework is a step ahead of the traditional approaches in detecting unusual activities, which shows a huge potential in automatically monitoring surveillance videos. Performing autonomous monitoring of CCTV, crime control and damage or destruction by a group of people or crowd can be identified and alarms may be triggered in unusual activities in streets or public places. The suggested GNN model improves accuracy by 4% for the Pedestrian 2 dataset and 12% for the Pedestrian 1 dataset compared to a few state-of-the-art techniques.     

Energy Efficient Clustering Protocol to Enhance Network Lifetime in Wireless Sensor Networks

In this paper, the energy conservation in the ununiform clustered network field is proposed. The fundamental reason behind the methodology is that in the process of CH election, nodes Competition Radius (CR) task is based on not just the space between nodes and their Residual Energy (RE), which is utilized in Energy-Aware Distributed Unequal Clustering (EADUC) protocol but also a third-degree factor, i.e., the nearby multi-hop node count. In contrast, a third-factor nearby nodes count is also used. This surrounding data is taken into account in the clustering feature to increase the network’s life span. The proposed method, known as Energy Conscious Scattered Asymmetric Clustering (ECSAC), self-controls the nodes’ energy utilization for equal allotment and un-equal delivery. Besides, extra attention is agreed to energy consumption in the communication process by applying a timeslot-based backtracking algorithm for increasing the network’s lifetime. The proposed methodology reduces the clustering overhead and node communication energy consumption to extend the network’s lifetime. Our suggested method’s results are investigated against the classical techniques using the lifetime of the network, RE, alive hop count and energy consumption during transmission as the performance metric.     

Using Object Detection Network for Malware Detection and Identification in Network Traffic Packets

In recent years, the number of exposed vulnerabilities has grown rapidly and more and more attacks occurred to intrude on the target computers using these vulnerabilities such as different malware. Malware detection has attracted more attention and still faces severe challenges. As malware detection based traditional machine learning relies on exports’ experience to design efficient features to distinguish different malware, it causes bottleneck on feature engineer and is also time-consuming to find efficient features. Due to its promising ability in automatically proposing and selecting significant features, deep learning has gradually become a research hotspot. In this paper, aiming to detect the malicious payload and identify their categories with high accuracy, we proposed a packet-based malicious payload detection and identification algorithm based on object detection deep learning network. A dataset of malicious payload on code execution vulnerability has been constructed under the Metasploit framework and used to evaluate the performance of the proposed malware detection and identification algorithm. The experimental results demonstrated that the proposed object detection network can efficiently find and identify malicious payloads with high accuracy.     

多特征双重匹配验证的驾驶员脸部融合检测

针对基于单一特征驾驶员脸部检测算法在检测精度和可靠性方面的局限性,提出了一种新颖的驾驶员脸部检测融合算法.首先采用改进的基于Haar-like特征的人脸检测算法在整幅图像上检测出可能存在的初始人脸区域,然后自适应地扩大初始人脸区域范围,并在此基础上利用基于肤色特征的方法在YCbCr空间上进行脸部的二次检测,最后根据定义的脸部区域重合度和人脸几何先验知识对驾驶员脸部区域进行双重匹配验证进而制定相应的定位规则对脸部进行融合检测.各种复杂路况下的实验结果证明了该算法的有效性.     

基于最近邻搜索耦合近邻损耗聚类的图像伪造检测算法

目的为了解决当前图像伪造检测算法在对图像进行伪造检测时,主要依靠全局搜索的方式来完成特征点匹配,导致其检测效率较低,且在对复杂伪造图像进行检测时,易出现检测精度不高和检测错误的不足。方法提出基于最近邻搜索耦合近邻损耗聚类的图像伪造检测算法。首先引入积分图像的方法,对图像进行预处理,借助Hessian矩阵行列式来提取特征点。利用特征点构建圆形区域,通过求取圆形区域内Haar小波响应获取特征点的特征描述符。然后通过特征描述符建立KD树索引,利用最近邻搜索方法代替SURF中全局搜索的方法,对SURF进行改进,完成特征点的匹配。最后,利用特征点间的近邻关系求取近邻函数值,通过近邻函数值对特征点进行聚类,完成图像的伪造检测。结果实验结果显示,与当前图像伪造检测算法相比,所提算法具有更高的检测效率以及更高的检测正确度。结论所提算法具备较高的检测精度,在印刷防伪与信息安全等领域具有较好的应用价值。     

A Model for Helmet-Wearing Detection of Non-Motor Drivers Based on YOLOv5s

Detecting non-motor drivers’ helmets has significant implications for traffic control. Currently, most helmet detection methods are susceptible to the complex background and need more accuracy and better robustness of small object detection, which are unsuitable for practical application scenarios. Therefore, this paper proposes a new helmet-wearing detection algorithm based on the You Only Look Once version 5 (YOLOv5). First, the Dilated convolution In Coordinate Attention (DICA) layer is added to the backbone network. DICA combines the coordinated attention mechanism with atrous convolution to replace the original convolution layer, which can increase the perceptual field of the network to get more contextual information. Also, it can reduce the network’s learning of unnecessary features in the background and get attention to small objects. Second, the Rebuild Bidirectional Feature Pyramid Network (Re-BiFPN) is used as a feature extraction network. Re-BiFPN uses cross-scale feature fusion to combine the semantic information features at the high level with the spatial information features at the bottom level, which facilitates the model to learn object features at different scales. Verified on the proposed “Helmet Wearing dataset for Non-motor Drivers (HWND),” the results show that the proposed model is superior to the current detection algorithms, with the mean average precision (mAP) of 94.3% under complex background.