Search

Abstract In the video compression industry, video compression tailored to machine vision tasks has recently emerged as a critical area of focus. Given the unique characteristics of machine vision, the current practice of directly employing conventional codecs reveals inefficiency, which requires compressing unnecessary regions. In this paper, we propose a framework that more aptly encodes video regions distinguished by machine vision to enhance coding efficiency. For that, the proposed framework consists of deep learning-based adaptive switch networks that guide the efficient coding tool for video encoding. Through the experiments, it is demonstrated that the proposed framework has superiority over the latest standardization project, video coding for machine benchmark, which achieves a Bjontegaard delta (BD)-rate gain of 5.91% on average and reaches up to a 19.51% BD-rate gain.

With the proliferation of the Internet, network complexities for both commercial and state organizations have significantly increased, leading to more sophisticated and harder-to-detect network attacks. This evolution poses substantial challenges for intrusion detection systems, threatening the cybersecurity of organizations and national infrastructure alike. Although numerous deep learning techniques such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), and graph neural networks (GNNs) have been applied to detect various network attacks, they face limitations due to the lack of standardized input data, affecting model accuracy and performance. This paper proposes a novel preprocessing method for flow data from network intrusion detection systems (NIDSs), enhancing the efficacy of a graph neural network model in malicious flow detection. Our approach initializes graph nodes with data derived from flow features and constructs graph edges through the analysis of IP relationships within the system. Additionally, we propose a new graph model based on the combination of the graph neural network (GCN) model and SAGEConv, a variant of the GraphSAGE model. The proposed model leverages the strengths while addressing the limitations encountered by the previous models. Evaluations on two IDS datasets, CICIDS-2017 and UNSW-NB15, demonstrate that our model outperforms existing methods, offering a significant advancement in the detection of network threats. This work not only addresses a critical gap in the standardization of input data for deep learning models in cybersecurity but also proposes a scalable solution for improving the intrusion detection accuracy.

Recently, compression is essential to reduce large traffic generated from video or image-based services such as OTT, IOT, smart city, and self-driving cars. The high data compression reduces data traffic, but causes information loss and data quality deterioration. In this paper, we shall propose a new efficient compression method which increases the compression ratio while maintaining the image quality by partitioning the image into 5 different areas each with the different quantization values according to their priority such as intensity, location, size and correlation. We exploit different compression ratio for each of the areas divided based on saliency map according to their visual priority order in order to fully reflect the visual reaction degree of human visual system. We shall show that our method dramatically increases data preservation and compression efficiency by compressing each area with the different quantization values while using the less memory.

A Software Defined Network (SDN) provides efficient network management by decoupling two planes: the control plane and the data plane. However, although SDN provides efficient network management, it also causes several critical vulnerabilities. In particular, the lack of memory for a flow table in the data plane can be exploited to conduct a flow table overflow attack. This paper proposes a history-based dynamic timeout scheme to mitigate the flow table overflow attack. The proposed scheme dynamically sets up both hard timeout and idle timeout based on statistical history for each flow, which can quickly remove attack flows from a flow table. Consequently, it can keep the occupancy of the flow table low and secure the robustness against the flow table overflow attack. The experiment results show that the proposed HSDT can mitigate the overflow attack with reasonable overhead by effectively evicting attack flow rules from the flow table while it has a minimal impact on the other normal flow rules and bandwidth.

Network slicing is a key technology in fifth-generation (5G) networks that allows network operators to create multiple logical networks over a shared physical infrastructure to meet the requirements of diverse use cases. Among core functions to implement network slicing, resource management and scaling are difficult challenges. Network operators must ensure the Service Level Agreement (SLA) requirements for latency, bandwidth, resources, etc for each network slice while utilizing the limited resources efficiently, i.e., optimal resource assignment and dynamic resource scaling for each network slice. Existing resource scaling approaches can be classified into reactive and proactive types. The former makes a resource scaling decision when the resource usage of virtual network functions (VNFs) exceeds a predefined threshold, and the latter forecasts the future resource usage of VNFs in network slices by utilizing classical statistical models or deep learning models. However, both have a trade-off between assurance and efficiency. For instance, the lower threshold in the reactive approach or more marginal prediction in the proactive approach can meet the requirements more certainly, but it may cause unnecessary resource wastage. To overcome the trade-off, we first propose a novel and efficient proactive resource forecasting algorithm. The proposed algorithm introduces an attention-based encoder-decoder model for multivariate time series forecasting to achieve high short-term and long-term prediction accuracies. It helps network slices be scaled up and down effectively and reduces the costs of SLA violations and resource overprovisioning. Using the attention mechanism, the model attends to every hidden state of the sequential input at every time step to select the most important time steps affecting the prediction results. We also designed an automated resource configuration mechanism responsible for monitoring resources and automatically adding or removing VNF instances of network slices, which helps network operators satisfy service requirements even when the traffic of end-user requests changes dynamically. Comprehensive experiments demonstrate that our proposed solution outperforms other solutions in terms of short-term and long-term predictions while reducing the cost of SLA violations and resource overprovisioning and enhancing the delay quality of network slices.

Image classification research is one of the fields continuously studied in the computer vision domain, and several related studies have been actively conducted until recently. However, a limit exists regarding the prediction performance of real-world datasets due to the data imbalance problem between classes. Data augmentation through artificial sample generation for minority classes is one of the methods used to overcome this limitation. Among the various oversampling methods, we propose the feature dictionary-based generative model for the oversampling method. Feature dictionaries are built through the pretrained feature extractor, and the proposed generative model synthesizes artificial samples based on the dictionary. Class-to-class balanced training can be conducted by fine-tuning the classifier as additional data for the minority class. We experiment by applying the proposed framework to the fashion dataset, which has an extreme class imbalance. The experimental results demonstrate that the proposed model achieved the highest top-1 performance on various public fashion datasets. In addition, we analyze the number of samples in the dictionary and test the effectiveness of the elements that comprise the proposed model using various ablation studies.

Various regulations are being devised and implemented to prevent the environmental pollution that is threatening mankind. The International Maritime Organization has strengthened regulations on sulfur, a notorious pollutant, to prevent sea pollution. In addition, the production of LNG fueled ships is increasing. Among various metals, 9% nickel steel is widely used in the shipbuilding industry because it is advantageous in terms of material strength and cryogenic impact toughness. Various studies are being carried out to predict and prevent its distortion, caused by welding, in the design. To predict welding distortion during flux core arc welding, this study found a way to refine the parameters constituting the Goldak welding heat source. The optimal heat source parameters were derived by using BOP experiments, cross-sectional analysis, finite element analysis and global optimization algorithm. When re-analyzed and verified based on the values, an error of up to 6.3% was found between simulation results and experimental values. The process was improved by clarifying the objective function and reducing the range of candidate welding efficiencies during global optimization and the process efficiency was also improved by reducing analysis time with a simplified model. Therefore, it is thought that this study can contribute to the productivity improvement of LNG storage containers, helping engineers apply it immediately in the industrial field.

Various international organizations and governments of many countries are making efforts to prevent environmental pollution, with the IMO (International Maritime Organization) reinforcing related regulations. With these regulations, equipment related to LNG-fueled ships, which have the greatest carbon dioxide reduction effect among eco-friendly ships, are expected to increase. Although the IGC code designates the materials that can be used for LNG containers, such as 304L stainless steel and 9% nickel steel, these materials have a tendency to deteriorate the tissue around the heat-affected zone due to excessive heat input. In this study, we analyzed the effect of brittle fracture in the weld zone and heat-affected zone after fiber laser welding and found that welding quality improved with control of the heat input. SVM discriminant analysis was applied to classify the groups in which brittle fracture and ductile fracture occurred. The shape of the penetration section, hardness in the welding zone and heat-affected zone, and fracture surface were selected as factors for discrimination; these values were determined under various welding conditions. With these data, we derived a regression model and multi-objective optimization algorithm to predict mechanical properties after welding, as well as the conditions necessary to prevent brittle fracture. Finally, the prediction models were verified, as the results of welding under the derived conditions were classified as ductile fracture group.

The International Maritime Organization has adopted the reduction of carbon dioxide emissions from ships as an important priority, and is continuously strengthening its regulations on marine air pollution. By 2035, it is expected that LNG-powered ships will account for more than 50% of the available ships. Accordingly, the demand for equipment related to LNG-fueled ships is expected to grow as well, requiring the development of a lot of equipment. However, the characteristics of LNG-powered ships mean that they require a high level of reliability and long history of operating reliably. Even when a product is developed, numerous demonstrations and quality assurance measures are needed to reach the technological level ship owners and customers require. Therefore, an optimization procedure to determine the welding quality for 9% Ni steel is necessary. In this study, the heat input criteria that induce brittle fracture characteristics were analyzed to optimize the flux core arc welding process for 9% Ni steel used in the manufacture of LNG storage tanks. We developed an optimization algorithm (Welding Current, Arc Voltage, Welding Speed) that can select a group of fracture conditions by examining the tendency of the tissue to brittle fracture due to excessive heat input among potential quality issues of cryogenic steel. Capable of selecting the range in which quality deterioration occurs, determining quality of a weld and avoiding the range in which toughness degradation occurs, through which a process to derive high quality 9% Ni welds is proposed.

In spite of protection mechanisms for Domain Name System (DNS), such as IP blacklist and DNS Firewall, DNS still has privacy issues in reality, since DNS is a plain-text protocol. Recently, to resolve this problem, an encrypted DNS, called DNS-over-HTTPS (DoH), has been developed, and is becoming more widespread. As the secured version of DNS, DoH guarantees privacy and security to prevent various attacks such as eavesdropping and manipulating DNS data by using the HTTPS protocol to encrypt the data between the DoH client and the DoH-based DNS resolver. DoH is one of the best security options for an enterprise network where more sensitive data protection is required. However, DoH may cause an unintended security breach, i.e., information leakage via malicious DoH tunneling. Since the DoH traffic is encrypted and indistinguishable from other HTTPS traffic, data hidden inside DoH packets can be easily leaked out of an enterprise network. Although some countermeasures to detect DoH tunneling attacks have been proposed, they still have limitations. Previous research used Supervised Machine Learning methods to detect DoH tunneling, which required a high volume of labeled data. In practice, collecting and labeling all of the data is an impossible task, especially in DoH, when all of the data are encrypted. Furthermore, Supervised Machine Learning methods rely heavily on human-engineered feature extraction, which makes classifying encrypted DoH traffic difficult. Furthermore, no previous research has mentioned a complete functional DoH detection applied to network infrastructure. Therefore, we propose a detection system for DoH tunneling attacks based on Transformer to detect a malicious DoH tunneling and build a fully functional DoH detection system that can be integrated with the security operation system of an enterprise network. The experiment results show a significant improvement compared with previous works.

Cloud computing is currently considered the most cost-effective platform for offering business and consumer IT services over the Internet. However, it is prone to new vulnerabilities. A new type of attack called an economic denial of sustainability (EDoS) attack exploits the pay-per-use model to scale up the resource usage over time to the extent that the cloud user has to pay for the unexpected usage charge. To prevent EDoS attacks, a few solutions have been proposed, including hard-threshold and machine learning-based solutions. Among them, long short-term memory (LSTM)-based solutions achieve much higher accuracy and false-alarm rates than hard-threshold and other machine learning-based solutions. However, LSTM requires a long sequence length of the input data, leading to a degraded performance owing to increases in the calculations, the detection time, and consuming a large number of computing resources of the defense system. We, therefore, propose a two-phase deep learning-based EDoS detection scheme that uses an LSTM model to detect each abnormal flow in network traffic; however, the LSTM model requires only a short sequence length of five of the input data. Thus, the proposed scheme can take advantage of the efficiency of the LSTM algorithm in detecting each abnormal flow in network traffic, while reducing the required sequence length of the input data. A comprehensive performance evaluation shows that our proposed scheme outperforms the existing solutions in terms of accuracy and resource consumption.

The article presents a study which explores the stretchable PPG sensor with light polarization for physical activity–permissible monitoring. It mentions that skin-attachable sensors, which represent the ultimate form of wearable electronic devices that ensure conformal contact with skin, suffer from motion artifact limitations owing to relative changes in position between the sensor and skin during physical activities.