Abstract With the widespread adoption of Software Defined Networking (SDN), detecting Distributed Denial Service (DDoS) attacks has become an urgent challenge in SDN maintenance and Security. Given diversity DDoS attack types, we face significant challenges. This paper proposes a model called ARSAE-QGRU, which is based on integrating attention mechanisms residual connections within stacked autoencoder for detection. By introducing into (SAE), effectively conveys more valuable information...

As a current research hotspot, graph convolution networks (GCNs) have provided new opportunities for tree species classification in multi-source remote sensing images. To solve the challenge of limited label information, model was proposed by using semi-supervised fusion method hyperspectral images (HSIs) and multispectral (MSIs). In model, graph-based attribute features pixel-based are fused to deepen correlation improve accuracy. Firstly, employs canonical analysis (CCA) maximize images,...

Hemp is an environmentally friendly porous fiber, and has good sound absorption properties, which can be used to reduce noise. In recent years, hemp-based composites have been widely studied, but most of them study the preparation methods materials, Therefore, this paper studies topic. Firstly, material specimens with different thickness diameter were prepared, then, standing wave tube method experiment was carried out, values coefficient measured analyzed, while a novel model using genetic...

The conventional TCP was designed to wired networks on the assumption that loss of data packet caused by network congestion, but in ad hoc networks, a large number due high BER(bits error rate), nodes mobility, etc. At consequence, congestion control mechanisms are not appropriate for wireless and extent reduces performance. In this paper, we analyze major factors affecting performance give several typical improved approaches, compare these different approaches. simulation results show...

Influence maximization of temporal social networks (IMT) is a problem that aims to find the most influential set nodes in network so their information can be widely spread. To solve IMT problem, we propose an influence algorithm based on improved K-shell method, namely (KT). The takes into account global and local structures networks. First, obtain kernel value <i>Ks</i> each node, scope, it layers according characteristic by improving method. Then, calculation method comprehensive degree...

Influence Maximization (IM) aims to select a seed set of size k in social network so that information can be spread most widely under specific propagation model through this nodes.However, existing studies on the IM problem focus static features, while neglecting features temporal networks.To bridge gap, we node reflected by their historical interaction behavior networks, i.e., attributes and self-similarity, incorporate them into influence maximization algorithm model.Firstly, propose...

The investigation of image deblurring techniques in dynamic scenes represents a prominent area research. Recently, deep learning technology has gained extensive traction within the field methodologies. However, such methods often suffer from limited inherent interconnections across various hierarchical levels, resulting inadequate receptive fields and suboptimal outcomes. In U-Net, more adaptable approach is employed, integrating diverse levels features effectively. Such design not only...

Current acoustic modeling methods face problems such as complex processes or inaccurate sound absorption coefficients, etc. Therefore, this paper studies the topic. Firstly, material samples were prepared, and standing wave tube method experiments conducted. Material data obtained, while a model using improved genetic algorithm neural network was subsequently proposed. Secondly, obtained from experiment analyzed; structure designed; training, verification, test all divided. In order to...

Deblurring methods in dynamic scenes are a challenging problem. Recently, significant progress has been made for image deblurring based on deep learning. However, these usually stack ordinary convolutional layers or increase convolution kernel size, resulting limited receptive fields, an unsatisfying effect, and heavy computational burden. Therefore, we propose improved U-Net (U-shaped Convolutional Neural Network) model to restore the blurred images. We first design structure, which mainly...

The relationship between users and items, which cannot be recovered by traditional techniques, can extracted the recommendation algorithm based on graph convolution network.The current simple linear combination of these algorithms may not sufficient to extract complex structure user interaction data.This paper presents a new approach address such issues, utilizing network association relations.The proposed mainly includes three modules: Embedding layer, forward propagation score prediction...

Abstract Intrusion detection represents an efficacious approach for addressing security concerns. However, given the substantial volume and high-dimensional nature of WLAN dataset features, existing methods exhibit limited effectiveness in feature extraction, thereby impacting classification performance. To address above problems, improved deep neural network (DNN) model intrusion was proposed. Firstly, activation function loss a single sparse autoencoders (SAE) were determined through...

For scanning electron microscopes with high resolution and a strong electric field, biomass materials under observation are prone to radiation damage from the beam. This results in blurred or non-viable images, which affect further of material microscopic morphology characterization. Restoring images their original sharpness is still challenging problem image processing. Traditional methods can't effectively separate context dependency texture information, effect enhancement deblurring,...

In order to solve the problems of low brightness contrast a color image, hiding large amount detail information, and deviation information in process image acquisition, an optimization method plane enhancement processing based on computer vision virtual reality is proposed. this method, input RGB converted into represented by HSI model, its adaptive adjusted improve overall image. For local three-dimensional Gaussian model perceived retinal neurons introduced illuminance estimation MSR...

The early stages of diabetic foot represent a critical treatment period, but patients show no obvious symptoms. Upon the development into ulcers, risk amputation exists for which costs are high. In this study, considering plantar pressure as an important physiological parameter foot, we proposed methods to assist diagnosis foot. Plantar images were collected and de-noised. An improved automatic regional division algorithm was proposed. Laplacian spectrum features extracted according maximum...

The mobility and intelligence of Agent make it have a very good application value, moreover, the study schemes on model for embedded mobile devices are rare, so article proposes an demand business applications, which is consist wireless compute environment model, general framework, core memory they designed as multi-system combining technology with Belief Desire Intention evolution thought, AUML specification used to draw design diagram during implementing system, finally server client...