• Journal of Information Processing Systems
• /
• v.11 no.1
• /
• pp.148-164
• /
• 2015

The use of mobile agents for collaborative processing in wireless sensor network has gained considerable attention. This is when mobile agents are used for data aggregation to exploit redundant and correlated data. The efficiency of agent-based data aggregation depends on the agent migration scheme. However, in general, most of the proposed schemes are centralized approach-based schemes where the sink node determines the migration paths for the agents before dispatching them in the sensor network. The main limitations with such schemes are that they need global network topology information for deriving the migration paths of the agents, which incurs additional communication overhead, since each node has a very limited communication range. In addition, a centralized approach does not provide fault tolerant and adaptive migration paths. In order to solve such problems, we have proposed a distributed approach-based scheme for determining the migration path of the agents where at each hop, the local information is used to decide the migration of the agents. In addition, we also propose a local repair mechanism for dealing with the faulty nodes. The simulation results show that the proposed scheme performs better than existing schemes in the presence of faulty nodes within the networks, and manages to report the aggregated data to the sink faster.

• Bulletin of the Korean Mathematical Society
• /
• v.51 no.2
• /
• pp.457-478
• /
• 2014

In this paper, a 3D autonomous system, which has only stable or non-hyperbolic equilibria but still generates chaos, is presented. This system is topologically non-equivalent to the original Lorenz system and all Lorenz-type systems. This motivates us to further study some of its dynamical behaviors, such as the local stability of equilibrium points, the Lyapunov exponent, the dissipativity, the chaotic waveform in time domain, the continuous frequency spectrum, the Poincar$\acute{e}$ map and the forming mechanism for compound structure of its special cases. Especially, with the help of the Project Method, its Hopf bifurcation of codimension one is in detailed formulated. Numerical simulation results not only examine the corresponding theoretical analytical results, but also show that this system possesses abundant and complex dynamical properties not solved theoretically, which need further attention.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.2
• /
• pp.786-799
• /
• 2021

Road damage detection is important for road maintenance. With the development of deep learning, more and more road damage detection methods have been proposed, such as Fast R-CNN, Faster R-CNN, Mask R-CNN and RetinaNet. However, because shallow and deep layers cannot be extracted at the same time, the existing methods do not perform well in detecting objects with fewer samples. In addition, these methods cannot obtain a highly accurate detecting bounding box. This paper presents a Multi-level Feature Pyramids method based on M2det. Because the feature layer has multi-scale and multi-level architecture, the feature layer containing more information and obvious features can be extracted. Moreover, an attention mechanism is used to improve the accuracy of local boundary boxes in the dataset. Experimental results show that the proposed method is better than the current state-of-the-art methods.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.7
• /
• pp.1907-1928
• /
• 2024

In recent years, multi-focus image fusion has emerged as a prominent area of research, with transformers gaining recognition in the field of image processing. Current approaches encounter challenges such as boundary artifacts, loss of detailed information, and inaccurate localization of focused regions, leading to suboptimal fusion outcomes necessitating subsequent post-processing interventions. To address these issues, this paper introduces a novel multi-focus image fusion technique leveraging the Swin Transformer architecture. This method integrates a frequency layer utilizing Wavelet Transform, enhancing performance in comparison to conventional Swin Transformer configurations. Additionally, to mitigate the deficiency of local detail information within the attention mechanism, Convolutional Neural Networks (CNN) are incorporated to enhance region recognition accuracy. Comparative evaluations of various fusion methods across three datasets were conducted in the paper. The experimental findings demonstrate that the proposed model outperformed existing techniques, yielding superior quality in the resultant fused images.

• Animal cells and systems
• /
• v.5 no.1
• /
• pp.31-35
• /
• 2001

Temperate-resident bats exhibit a circadian cycle of torpor and arousal In summer. The physiological role and selective advantage of torpor as an energy saving mechanism have been received much attention by hibernation biologists. However, despite the significance of the recovering euthermic function, the arousal process and mechanism in these animals have been poorly addressed. In this study, we investigated thermogenic and motor activities of a local bat species Murina leucogaster during arousal by simultaneously examining oxygen consumption rate, body temperature ($T_b$) and pectoral electromyography (EMG). We found that Tb of the torpid bats (12-14$^{\circ}C$) was augmented slowly by nonshivering mechanism during the initial awakening phase. The pectoral shivering, gauged by EMG activity, occurred between 17$^{\circ}C$ and 38$^{\circ}C$. Over this Tb range of shivering, heat was produced at a rate of 0.145 kcal $kg^{-1}\;min^{-1}$ to raise 1$^{\circ}C$ $T_b$ per min. Shivering was most intensive at 30-35$^{\circ}C$ where both EMG amplitude and spike frequency were the highest. Activation of the pectoral myofibers seemed to be controlled in a manner that motor units were recruited from smaller to larger sizes, with greater synchronization, as muscle shivering became intensive with increasing $T_b$.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.42 no.2
• /
• pp.523-535
• /
• 2017

Local intelligent services aim at controlling local services such as cooling or lightening services in a certain local area, using Internet-of-Things (IoT) sensor data in the area. As the IoT paradigm has evolved, local intelligent services have gained increasing attention. However, most of the local intelligent service mechanism proposed so far do not directly take the users' presence and service preference information into account for controlling local services. This study proposes an individual presence-and-preference-based local service system (IPP-LISS). We present a intelligent service control algorithm and implement a prototype system of IPP-LISS. Typically, the intelligence part of IPP-LISS including the prediction models, is generated on remote server in the cloud because of their compute-intense aspect. However, this can cause huge data traffic between IoT devices and servers in the cloud. The emerging mobile edge computing technology will be a promising solution of this challenge of IPP-LISS. In this paper, we implement IPP-LISS in the cloud, and then, based on the implementation result, we discuss applying the mobile edge computing technology to the IPP-LISS application.

• The Journal of Korean Orthopaedic Ultrasound Society
• /
• v.4 no.2
• /
• pp.101-110
• /
• 2011

To describe the background, mechanism, clinical results and complications of prolotheapy based on the literature review. Prolotherapy is a minimally invasive injection-based treatment of chronic musculoskeletal pain, including ligament and joint laxity. The mechanism of this injection-based technique is to initiate a local inflammatory response with resultant tissue healing. The used proliferants are classified by bio-mechanism to act in three different ways as osmotic, irritants, and chemotatics. The most commonly used proliferant is hyperosmolar (10~25%) dextrose to act by osmotic rupture of cells. High resolution ultrasound imaging of musculoskeletal structure provide a more accurate diagnosis. Also ultrasound-guided intervention provides a more high efficacy and low rate of complications. The most common complication is local pain at the injected site, that is self-limited and good responsive to anti-inflammatory agents. Other complications are rare. It is reported that prolotherapy appears safe when performed by an experienced clinician. Prolotherapy has grown in popularity and has received significant recent attention. However there are limited evidence-based data supporting the indication and efficacy of prolotherapy in the treatment of chronic musculoskeletal pain or soft tissue injuries. Future studies are necessary to determine whether prolotherapy can play an independent and definitive role in a treatment for chronic musculoskeletal pain.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.9
• /
• pp.4242-4263
• /
• 2017

Multi-User Multiple-Input Multiple-Output (MU-MIMO) technology has recently attracted significant attention from academia and industry because of it is increasingly important role in improving networks' capacity and data rate. Moreover, MU-MIMO systems for the Fifth Generation (5G) have already been researched. High Quality of Service (QoS) and efficient operations at the Medium Access Control (MAC) layer have become key requirements. In this paper, we propose a downlink MU-MIMO MAC protocol based on adaptive Channel State Information (CSI) feedback (called MMM-A) for Wireless Local Area Networks (WLANs). A modified CSMA/CA mechanism using new frame formats is adopted in the proposed protocol. Specifically, the CSI is exchanged between stations (STAs) in an adaptive way, and a packet selection strategy which can guarantee a fairer QoS for scenarios with differentiated traffic is also included in the MMM-A protocol. We then derive the expressions of the throughput and access delay, and analyze the performance of the protocol. It is easy to find that the MMM-A protocol outperforms the commonly used protocols in terms of the saturated throughput and access delay through simulation and analysis results.

• Journal of Arbitration Studies
• /
• v.23 no.4
• /
• pp.29-53
• /
• 2013

Chinese ISD has been changed a lot since the reformation policy in 1978 and it is expected that China will present a changed attitude toward its advantage as its industrialization continues to advance. This study generally examines the ISD in BIT and also considers not only the attitude of China with regard to ISD but also the changes on the Chinese side. Moreover, this study determines the areas on which the Chinese government focuses. In order to conduct this study, the author attempts to classify the attitudes on ISD into chronical change and treaty powers based on the analysis of BIT. In addition, the paper examines the main contents of ISD in BIT which previously involved an agreement such as arbitral institution, arbitral range, counter-measures of local country, standard for admitting the nationality of corporate investors, and recognition and enforcement of arbitral award. Based on analysis, this paper mentions matters that require attention and caution in the Korea-China FTA as regards investment negotiation, and also suggests instructions for investors who may face dispute with the Chinese government.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.12
• /
• pp.3868-3888
• /
• 2022

A widely used social networking service like Twitter has the ability to disseminate information to large groups of people even during a pandemic. At the same time, it is a convenient medium to share irrelevant and unverified information online and poses a potential threat to society. In this research, conventional machine learning algorithms are analyzed to classify the data as either non-rumor data or rumor data. Machine learning techniques have limited tuning capability and make decisions based on their learning. To tackle this problem the authors propose a deep learning-based Rumor Detection Neural Network model to predict the rumor tweet in real-world events. This model comprises three layers, AttCNN layer is used to extract local and position invariant features from the data, AttBi-LSTM layer to extract important semantic or contextual information and HPOOL to combine the down sampling patches of the input feature maps from the average and maximum pooling layers. A dataset from Kaggle and ground dataset #gaja are used to train the proposed Rumor Detection Neural Network to determine the veracity of the rumor. The experimental results of the RDNN Classifier demonstrate an accuracy of 93.24% and 95.41% in identifying rumor tweets in real-time events.