A low complexity modulation format identification based on unsupervised Kmeans clustering algorithm and cluster validity index is proposed. Simulations show the scheme's 100% identification accuracy and stable per...
详细信息
Security is the main challenge in Internet of Things (IoT) systems. The devices on the IoT networks are very heterogeneous, many of them have limited resources, and they are connected globally, which makes the IoT muc...
详细信息
Security is the main challenge in Internet of Things (IoT) systems. The devices on the IoT networks are very heterogeneous, many of them have limited resources, and they are connected globally, which makes the IoT much more challenging to secure than other types of networks. Denial of service (DoS) is the most popular method used to attack IoT networks, either by flooding services or crashing services. Intrusion detection system (IDS) is one of countermeasures for DoS attack. Unfortunately, the existing IDSs are still suffering from detection accuracy problem due to difficulty of recognizing features of the DoS attacks. Thus, we need to determine specific features that representing well the traffic attacks, so the IDS will be able to distinguish normal traffic from the attacks. In this work, we investigate ping flood attack pattern recognition on IoT networks. Experiments were conducted using wireless communication with three different scenarios: normal traffic, attack traffic, and combined normal-attack traffic. Each scenario created an associated dataset. The datasets were then grouped into two clusters: normal and attack. The K-Means algorithm was used to produce the clustering results. The average number of packets in the attack cluster was 95 931 packets, and the average in the normal cluster was 4,068 packets. The accuracy level of the clustering results was calculated using a confusion matrix. The accuracy of the clustering using the implemented K-Means algorithm was 99.94%. The rates from the confusion matrix were true negative (98.62%), true positive (100.00%), false negative (0.00%), and false positive (1.38%).
As of late, with the progression of AI and man-made brainpower, there has been a developing spotlight on versatile e-learning. As all ways to deal with e-learning lose their allure and the level of online courses buil...
详细信息
As of late, with the progression of AI and man-made brainpower, there has been a developing spotlight on versatile e-learning. As all ways to deal with e-learning lose their allure and the level of online courses builds, they move towards more customized versatile learning so as to collaborate with students and achieve better learning results. The schools focus on the examination, mindfulness, and arranging techniques that infuse innovation into the vision and educational program. E-learning issues are a standard examination issue for us all. The motivation behind this research analysis is to separate the potential outcomes of assessing e-learning models utilizing AI strategies such as Supervised, Semi Supervised, Reinforced Learning advances by investigating upsides and downsides of various methods organization. The literature review methodology is to review the cross sectional impacts of e-learning and Machine learning algorithms from existing literatures from the year 1993 to 2020 and to assess the essentialness of e-learning features to optimize the e-learning models with available Machine learning techniques from peer-inspected journals, capable destinations, and books. Second, it legitimizes the chances of e-learning structures introduction, and changes demonstrated through AI and Machine Learning algorithms. This examination assists in providing helpful new highlights to analysts, researchers and academicians. It gives an exhaustive structure of existing e-learning frameworks for the most recent innovations identified with learning framework capacities and learning tasks to envision ML research openings in appropriate spaces. The survey paper identifies and demonstrates the important role of different types of e-learning features such as Individual pertinent feature, Course pertinent feature, Context pertinent feature and Technology pertinent feature in framework performance tuning. The performance of Machine Learning algorithms to optimize the features of E-
Video over-segmentation into supervoxels is an important pre-processing technique for many computer vision tasks. Videos are an order of magnitude larger than images. Most existing methods for generating supervovels a...
详细信息
Video over-segmentation into supervoxels is an important pre-processing technique for many computer vision tasks. Videos are an order of magnitude larger than images. Most existing methods for generating supervovels are either memory- or time-inefficient, which limits their application in subsequent video processing tasks. In this paper, we present an anisotropic supervoxel method, which is memory-efficient and can be executed on the graphics processing unit (GPU). Therefore, our algorithm achieves good balance among segmentation quality, memory usage and processing time. In order to provide accurate segmentation for moving objects in video, we use the optical flow information to design a brand new non-Euclidean metric to calculate the anisotropic distances between seeds and voxels. To efficiently compute the anisotropic metric, we adjust the classic jump flooding algorithm (which is designed for parallel execution on the GPU) to generate anisotropic Voronoi tessellation in the combined color and spatio-temporal space. We evaluate our method and the representative supervoxel algorithms for their capability on segmentation performance, computation speed and memory efficiency. We also apply supervoxel results to the application of foreground propagation in videos to test the performance on solving practical problems. Experiments show that our algorithm is much faster than the existing methods, and achieves good balance on segmentation quality and efficiency.
The current research on cooperative localization of mobile cluster networks mainly focus on 2D networks with high-quality wireless environment. This study considers the problem in 3D mobile networks, a fault-tolerant ...
详细信息
The current research on cooperative localization of mobile cluster networks mainly focus on 2D networks with high-quality wireless environment. This study considers the problem in 3D mobile networks, a fault-tolerant cooperative localization algorithm based on two-layer filter Multidimensional Scaling (Filter-MDS) is proposed. First, the non-line-of-sight (NLOS) error in the wireless environment is expanded, and the concept of random error environment is proposed. Secondly, a location framework suitable for 3D mobile networks is constructed and the Cramer-Rao lower bound (CRLB) of 3D location in random error environment is derived. Filters based on multi-time motion state and geometric configuration are introduced into the location framework to construct Filter-MDS. Simulation experiments show that Filter-MDS can effectively identify the abnormal errors in range measurement and improve cooperative localization accuracy of mobile cluster network in random error environment.
In order to achieve intelligent classification of bearing faults, after comparing a large number of mechanical fault signal features, this paper proposes a bearing intelligent diagnosis algorithm based on vibration po...
详细信息
Uncertainty and distributed nature inherently exist in big data environment. Distributed fuzzy neural network (D-FNN) that not only employs fuzzy logics to alleviate the uncertainty problem but also deal with data in ...
详细信息
Uncertainty and distributed nature inherently exist in big data environment. Distributed fuzzy neural network (D-FNN) that not only employs fuzzy logics to alleviate the uncertainty problem but also deal with data in a distributed manner, is effective and crucial for big data. Existing D-FNNs always avoided consensus for their antecedent layer due to computational difficulty. Hence such D-FNNs are not really distributed since a single model can not be agreed by multiple agents. This article proposes a true D-FNN model to handle the uncertainty and distributed challenges in the big data environment. The proposed D-FNN model considers consensus for both the antecedent and consequent layers. A novel consensus learning, which involves a distributed structure learning and a distributed parameter learning, is proposed to handle the D-FNN model. The proposed consensus learning algorithm is built on the well-known alternating direction method of multipliers, which does not exchange local data among agents. The major contribution of this paper is to propose the true D-FNN model for the big data and the novel consensus learning algorithm for this D-FNN model. Simulation results on popular datasets demonstrate the superiority and effectiveness of the proposed D-FNN model and consensus learning algorithm.
Field observation systems are mainly deployed in the harsh natural environment. These systems principally focus on observation and study within the station currently, which leads to problems such as the inability to f...
详细信息
Field observation systems are mainly deployed in the harsh natural environment. These systems principally focus on observation and study within the station currently, which leads to problems such as the inability to form combined network observation and quite challenging to answer the scientific questions of wider regions and scales. To form Field Observation Instruments Networks (FOINs) and accelerate the general automation rate as well as in real-time data exchange in field observation, a multi-objective decision-making mehod named Entropy-based TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) clustering routing algorithm (ETC) for FOIN is proposed in this paper. The ETC algorithm can select the optimal cluster head (Optimal-CH) through multi-objective decision-making and mainly solves the problem that some existing multi-objective optimization algorithms cannot dynamically and objectively allocate weights. The ETC algorithm was compared with some latest work and similar kinds of work from network lifespan, the number of CH and energy consumption in the Matlab simulations experiments. The result shows that the ETC algorithm performs well, enhancing energy conservation and extending the existence of FOIN.
Distribution topology is oftentimes changed to cope with the development of local power load. Therefore, connectivity verification has become a critical task for optimal grid operation. In this article, a novel cloud-...
详细信息
Distribution topology is oftentimes changed to cope with the development of local power load. Therefore, connectivity verification has become a critical task for optimal grid operation. In this article, a novel cloud-edge collaboration approach is presented to identify outlier users and correct connections. In this article, based on the smart meter voltage analytics, an affinity propagation clustering-based local outlier factor (AP-LOF) algorithm is proposed for the voltage outlier identification and verification of the edge transformer. Compared to traditional methods, it can effectively identify the outlier user groups with high internal voltage correlation. Besides, a recommendation mechanism is developed in the cloud center, which repositions the identified outlier users by coordinating the information exchange between the edge transformers and the cloud center. Numerical tests are conducted using the actual smart meter voltage data. The results show that the proposed AP-LOF algorithm exhibits a better performance, which is suitable for the identification of various outlier users. Compared to a centralized architecture, 66% savings in calculation time is achieved by the cloud-edge collaboration approach. It further demonstrates the effectiveness and practicability of the proposed method in terms of identification accuracy and verification efficiency.
In this paper, we propose a family of graph partition similarity measures that take the topology of the graph into account. These graph-aware measures are alternatives to using set partition similarity measures that a...
详细信息
In this paper, we propose a family of graph partition similarity measures that take the topology of the graph into account. These graph-aware measures are alternatives to using set partition similarity measures that are not specifically designed for graphs. The two types of measures, graph-aware and set partition measures, are shown to have opposite behaviors with respect to resolution issues and provide complementary information necessary to compare graph partitions.
暂无评论