• Journal of Communications and Networks
• /
• v.5 no.3
• /
• pp.275-281
• /
• 2003

Random Early Detection (RED) [1] is an active queue management scheme which has been deployed extensively to reduce packet loss during congestion. Although RED can improve loss rates, its performance depends severely on the tuning of its operating parameters. The idea of adaptively varying RED parameters to suit the network conditions has been investigated in [2], where the maximum packet dropping probability $max_p$ has been varied. This paper focuses on adaptively varying the queue weight $\omega_q$ in conjunction with $max_p$ to improve the performance. We propose two algorithms viz., $\omega_q$-thresh and $\omega_q$-ewma to adaptively vary $\omega_q$. The performance is measured in terms of the packet loss percentage, link utilization and stability of the instantaneous queue length. We demonstrate that varying $\omega_q$ and $max_p$ together results in an overall improvement in loss percentage and queue stability, while maintaining the same link utilization. We also show that $max_p$ has a greater influence on loss percentage and queue stability as compared to $\omega_q$, and that varying $\omega_q$ has a positive influence on link utilization.

• The Bulletin of The Korean Astronomical Society
• /
• v.45 no.1
• /
• pp.42.3-42.3
• /
• 2020

We investigate the observational properties of Wolf-Rayet stars, suggest the constraint of their mass-loss rate and apply our results to the observed progenitor candidates of Type Ib/Ic supernovae (iPTF13bvn and SN 2017ein). For this purpose, we adopt the WR star models with various mass-loss rates and wind terminal velocities. We obtain the high resolution spectra of those models at the pre-supernova phase using the radiative transfer code CMFGEN. We verify the optically faint property of SN Ic progenitors and show that the optical faintness is mainly originated by the high effective temperature at the photosphere. We also show that a simple analytic model for WR winds using a constant opacity can roughly predict the photospheric parameters. We show that the change of the mass-loss rate and the terminal wind velocity critically affects the optical luminosity. We find the optical luminosities of SN Ic progenitor models with our fiducial mass-loss rate prescription are fainter than the detection limits. We also suggest the mass-loss rate of WR stars may not exceed 2 times of our fiducial value by comparing our predictions with the detection limit of SN Ib/Ic progenitors. The directly observed progenitor candidate of iPTF13bvn can be explained by our SN Ib progenitor models. We find that the SN 2017ein progenitor candidate is too bright and too blue to be a SN Ic progenitor.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.18 no.5
• /
• pp.538-547
• /
• 2015

The goal of this study is to develop an algorithm to propose optimal deployment of detection sensor nodes in the target area, based on a performance surface, which represents detection performance of active and passive acoustic sonar systems. The performance surface of the active detection system is calculated from the azimuthal average of maximum detection ranges, which is estimated with a transmission loss and a reverberation level predicted using ray-based theories. The performance surface of the passive system is calculated using the transmission loss model based on a parabolic equation. The optimization of deployment configurations is then performed by a hybrid method of a virtual force algorithm and a particle swarm optimization. Finally, the effectiveness of deployment configurations is analyzed and discussed with the simulation results obtained using the algorithm proposed in this paper.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.63-75
• /
• 2022

The raw data collected by structural health monitoring (SHM) systems may suffer multiple patterns of anomalies, which pose a significant barrier for an automatic and accurate structural condition assessment. Therefore, the detection and classification of these anomalies is an essential pre-processing step for SHM systems. However, the heterogeneous data patterns, scarce anomalous samples and severe class imbalance make data anomaly detection difficult. In this regard, this study proposes a convolutional neural network-based data anomaly detection method. The time and frequency domains data are transferred as images and used as the input of the neural network for training. ResNet18 is adopted as the feature extractor to avoid training with massive labelled data. In addition, the focal loss function is adopted to soften the class imbalance-induced classification bias. The effectiveness of the proposed method is validated using acceleration data collected in a long-span cable-stayed bridge. The proposed approach detects and classifies data anomalies with high accuracy.

• Nuclear Engineering and Technology
• /
• v.56 no.4
• /
• pp.1284-1295
• /
• 2024

Due to increasing operational security demands, digital and intelligent condition monitoring of nuclear power plants is becoming more significant. However, establishing an accurate and effective anomaly detection model is still challenging. This is mainly because of data characteristics of nuclear power data, including the lack of clear class labels combined with frequent interference from outliers and anomalies. In this paper, we introduce a Transformer-based unsupervised model for anomaly detection of nuclear power data, a modified loss function based on the maximum correntropy criterion (MCC) is applied in the model training to improve the robustness. Experimental results on simulation datasets demonstrate that the proposed Trans-MCC model achieves equivalent or superior detection performance to the baseline models, and the use of the MCC loss function is proven can obviously alleviate the negative effect of outliers and anomalies in the training procedure, the F1 score is improved by up to 0.31 compared to Trans-MSE on a specific dataset. Further studies on genuine nuclear power data have verified the model's capability to detect anomalies at an earlier stage, which is significant to condition monitoring.

• Journal of Communications and Networks
• /
• v.7 no.3
• /
• pp.337-352
• /
• 2005

In this paper, a fuzzy logic implementation of the random early detection (RED) mechanism [1] is presented. The main objective of the proposed fuzzy controller is to reduce the loss probability of the RED mechanism without any change in channel utilization. Based on previous studies, it is clear that the performance of RED algorithm is extremely related to the traffic load as well as to its parameters setting. Using fuzzy logic capabilities, we try to dynamically tune the loss probability of the RED gateway. To achieve this goal, a two-input-single-output fuzzy controller is used. To achieve a low packet loss probability, the proposed fuzzy controller is responsible to control the $max_{p}$ parameter of the RED gateway. The inputs of the proposed fuzzy controller are 1) the difference between average queue size and a target point, and 2) the difference between the estimated value of incoming data rate and the target link capacity. To evaluate the performance of the proposed fuzzy mechanism, several trials with file transfer protocol (FTP) and burst traffic were performed. In this study, the ns-2 simulator [2] has been used to generate the experimental data. All simulation results indicate that the proposed fuzzy mechanism out performs remarkably both the traditional RED and Adaptive RED (ARED) mechanisms [3]-[5].

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.468-475
• /
• 2020

A good acoustic word embedding that can well express the characteristics of word plays an important role in wake-up word detection (WWD). However, the representation ability of acoustic word embedding may be weakened due to various types of environmental noise occurred in the place where WWD works, causing performance degradation. In this paper, we proposed triplet loss based Domain Adversarial Training (tDAT) mitigating environmental factors that can affect acoustic word embedding. Through experiments in noisy environments, we verified that the proposed method effectively improves the conventional DAT approach, and checked its scalability by combining with other method proposed for robust WWD.

• Biotechnology and Bioprocess Engineering:BBE
• /
• v.11 no.4
• /
• pp.368-371
• /
• 2006

The standard procedure outlined by the United States Environmental Protection Agency (US EPA) in Method 1623 for analyzing Giardia lamblia cysts and Cryptosporidium parvum oocysts in water samples consists of filtration, elution, centrifugal concentration, immunomagnetic separation (IMS), and immunofluorescence assay (IFA) followed by microscopic examination. In this study, the extent of (oo)cyst loss in each step of this procedure was evaluated by comparing recovery yields in segmented analyses: (i) IMS + IFA, (ii) concentration + IMS + IFA, and (iii) filtration/elution + concentration + IMS + IFA. The complete (oo)cyst recovery by the full procedure was $52{\sim}57%$. The (oo) cyst loss in the IMS step was only $0{\sim}6%$, implying that IMS is a fairly reliable method for (oo)cyst purification. Centrifugal concentration of the eluted sample and pellet collection before IMS resulted in a loss of $8{\sim}14%$ of the (oo)cysts. The largest (oo)cyst loss occurred in the elution step, with $68{\sim}71%$ of the total loss. The permeated loss of (oo)cysts was negligible during filtration of the water sample with a $1.0-{\mu}m$ pore polyethersulfone (PES) capsule. These results demonstrated that the largest fraction of (oo)cyst loss in this procedure occurred due to poor elution from the filter matrix. Improvements in the elution methodology are therefore required to enhance the overall recovery yield and the reliability of the detection of these parasitic protozoa.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.12
• /
• pp.2167-2172
• /
• 2008

A large scale embedded generator or a micro-grid is usually interconnected to the grid through the dedicated line. Immediately after the removal of the grid supply, it must be disconnected and remain disconnected until the normal grid supplies are restored. This paper proposes an islanding detection algorithm based on complex power at the intertie. Complex power at the intertie delivered to the grid has a value depending on power flow during normal operation. When loss of mains occurs, complex power at the intertie moves to a value depending on the parameters of the dedicated line. The performance of the proposed algorithm was investigated under various operating and fault conditions. The proposed algorithm can detect an islanding operation even if no power is exchanged prior to loss of mains and discriminate the fault condition from islanding.

• ETRI Journal
• /
• v.40 no.1
• /
• pp.133-145
• /
• 2018

Robustness, payload, and imperceptibility of audio watermarking algorithms are contradictory design issues with high-level security of the watermark. In this study, the major issue in achieving high payload along with adequate robustness against challenging signal-processing attacks is addressed. Moreover, a security code has been strategically used for secure transmission of data, providing tamper detection at the receiver end. The high watermark payload in this work has been achieved by using the complementary features of third-level detailed coefficients of discrete wavelet transform where the human auditory system is not sensitive to alterations in the audio signal. To counter the watermark loss under challenging attacks at high payload, Daubechies wavelets that have an orthogonal property and provide smoother frequencies have been used, which can protect the data from loss under signal-processing attacks. Experimental results indicate that the proposed algorithm has demonstrated adequate robustness against signal processing attacks at 4,884.1 bps. Among the evaluators, 87% have rated the proposed algorithm to be remarkable in terms of transparency.