• Journal of KIISE:Information Networking
• /
• v.31 no.1
• /
• pp.37-52
• /
• 2004

Broadcast has been often used to disseminate the frequently requested data efficiently to a large volume of mobile units over a single or multiple channels. Since the mobile units have limited battery power, the minimization of the access time for the broadcast data is an important problem. There have been many research efforts that focus on the improvement if the broadcast techniques by providing indexes on the broadcast data. In this paper, we studied an efficient index allocation method for the broadcast data over multiple physical channels, which cannot be coalesced into a single high bandwidth channel. Previously proposed index allocation techniques either require the equal size of index and data or have a performance degradation problem when the number of given physical channels is not enough. These two problems will result in the increased average access time for the broadcast data. To cope with these problems, we propose an efficient tree- structured index allocation method for the broadcast data with different access frequencies over multiple physical channels. Our method minimizes the average access time for the broadcast data by broadcasting the hot data and their indexes more often than the less hot data and their indexes. We present an in-0e0th experimental and theoretical analysis of our method by comparing it with other similar techniques. Our performance analysis shows that it significantly decrease the average access time for the broadcast data over existing methods.

• KIISE Transactions on Computing Practices
• /
• v.23 no.9
• /
• pp.517-534
• /
• 2017

This paper proposes an Internet of Things (IoT) middleware called Middleware for Cooperative Interaction of Things (MinT). MinT supports a fully distributed IoT environment in which IoT devices directly connect to peripheral devices, easily constructing a local or global network and sharing their data in an energy efficient manner. MinT provides a sensor abstract layer, a system layer and an interaction layer. These layers enable integrated sensing device operations, efficient resource management, and interconnection between peripheral IoT devices. In addition, MinT provides a high-level API, allowing easy development of IoT devices by developers. We aim to enhance the energy efficiency and performance of IoT devices through the performance improvements offered by MinT resource management and request processing. The experimental results show that the average request rate increased by 25% compared to existing middlewares, average response times decreased by 90% when resource management was used, and power consumption decreased by up to 68%. Finally, the proposed platform can reduce the latency and power consumption of IoT devices.

• The KIPS Transactions:PartA
• /
• v.16A no.6
• /
• pp.481-488
• /
• 2009

GIS is used to isolate large power electrical equipment using SF6 gas. While GIS has simple structure, it has few break down, relatively high reliability. But it is hard to check up faults for reason of pressure. Faults of GIS should have a ripple effect on community and be hard to recovery. Consequently, GIS imports a preventive diagnostic system to find internal faults in advance. It is most important that reliability on the GIS preventive diagnostic system, because it estimates abnormality of system by analysis result of collected data. But, exist system which used central data management is low efficiency, and hard to guarantee timeliness and accuracy of data. To guarantee timeliness and accuracy, the GIS preventive diagnostic system needs accordingly to use a real-time middleware. So, in this paper, to improve reliability of the GIS preventive diagnostic system, we use a middleware based on TMO for guaranteeing timeliness of real-time distributed computing. And we propose an improved GIS preventive diagnostic system applying data acquisition, monitoring and control methods based on the TMO model. The presented system uses the Communication Control Unit(CCU) for distributed data handling which is supported by TMO. CCU can improve performance of the GIS preventive diagnostic system by guaranteeing timeliness of data handling process and increasing reliability of data through the TMO middleware. And, it has designed to take full charge of overload on a data acquisition task had been processed in an exist server. So, it could reduce overload of the server and apply distribution environment from now. Therefore, the proposed system can improve performance and reliability of the GIS preventive diagnostic system and contribute to stable operation of GIS.

• International Journal of Advanced Culture Technology
• /
• v.5 no.1
• /
• pp.58-63
• /
• 2017

Emergencies and disasters can happen any time without any warning, and things can change and escalate very quickly, and often it is swift and decisive actions that make all the difference. It is a responsibility of the building facility management to ensure that a proven evacuation plan in place to cover various worst scenario to handled automatically inside the facility. To mapping out optimal safe escape routes is a straightforward undertaking, but does not necessarily guarantee residents the highest level of protection. The emergency evacuation navigation approach is a state-of-the-art that designed to evacuate human livings during an emergencies based on real-time decisions using live sensory data with pre-defined optimum path finding algorithm. The poor decision on causalities and guidance may apparently end the evacuation process and cannot then be remedied. This paper propose a cloud connected emergency evacuation system model to react dynamically to changes in the environment in emergency for safest emergency evacuation using IoT based emergency exit sign system. In the previous researches shows that the performance of optimal routing algorithms for evacuation purposes are more sensitive to the initial distribution of evacuees, the occupancy levels, and the type and level of emergency situations. The heuristic-based evacuees routing algorithms have a problem with the choice of certain parameters which causes evacuation process in real-time. Therefore, this paper proposes an evacuee routing algorithm that optimizes evacuation by making using high computational power of cloud servers. The proposed algorithm is evaluated via a cloud-based simulator with different "simulated casualties" are then re-routed using a Dijkstra's algorithm to obtain new safe emergency evacuation paths against guiding evacuees with a predetermined routing algorithm for them to emergency exits. The performance of proposed approach can be iterated as long as corrective action is still possible and give safe evacuation paths and dynamically configure the emergency exit signs to react for real-time instantaneous safe evacuation guidance.

• Journal of KIISE:Information Networking
• /
• v.30 no.5
• /
• pp.631-640
• /
• 2003

Because of multi-hop wireless communication, P2P applications in ad hoc networks suffer poor performance. We Propose neighbor caching strategy to overcome this shortcoming and show it is more efficient than self caching that nodes store data in theirs own cache individually. A node can extend its caching storage instantaneously with neighbor caching by borrowing the storage from idle neighbors, so overcome multi-hop wireless communications with data source long distance away from itself. We also present the ranking based prediction that selects the most appropriate neighbor which data can be stored in. The node that uses the ranking based prediction can select the neighbor that has high possibility to keep data for a long time and avoid caching the low ranked data. Therefore the ranking based prediction improves the throughput of neighbor caching. In the simulation results, we observe that neighbor caching has better performance, as large as network size, as long as idle time, and as small as cache size. We also show the ranking based prediction is an adaptive algorithm that adjusts times of data movement into the neighbor, so makes neighbor caching flexible according to the idleness of nodes

• KIISE Transactions on Computing Practices
• /
• v.21 no.7
• /
• pp.506-511
• /
• 2015

IOV is a technology that supports one or more virtual desktops, and can share a single physical device. In general, the virtual desktop uses the virtual IO devices which are provided by virtualization SW, using SW emulation technology. Virtual desktops that use the IO devices based on SW emulation have a problem in which service quality and performance are declining. Also, they cannot support the high-end application operations such as 3D-based CAD and game applications. In this paper, we propose a physical network separation system using Virtual Desktop Service based on HW direct assignments to overcome these problems. The proposed system provides independent desktops that are used to access the intranet or internet using server virtualization technology in a physical desktop computer for the user. In addition, this system can also support a network separation without network performance degradation caused by inspection of the network packet for logical network separations and additional installations of the desktop for physical network separations.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.1
• /
• pp.1-10
• /
• 2010

Recently heterogeneous storage system such as hybrid hard disk drive (H-HDD) combining flash memory and magnetic disk is launched, according as the read performance of NAND flash memory is enhanced as similar to that of hard disk drive (HDD) and the power consumption of NAND flash memory is reduced less than that of HDD. However, the read and write operations of NAND flash memory are slower than those of rotational disk. Besides, serious overheads are incurred on CPU and main memory in the case that intensive write requests to flash memory are repeatedly occurred. In this paper, we propose the Least Frequently Used-Hot scheme that replaces the data blocks whose reference frequency of read operation is low and update frequency of write operation is high, and the data flushing scheme that rearranges the data blocks into the multi-zone of the rotation disk. Experimental results show that the execution time of the proposed method is 38% faster than those of conventional LRU and LFU block replacement schemes in I/O performance aspect and the proposed method increases the life span of Non-Volatile Cache 40% higher than those of conventional LRU, LFU, FIFO block replacement schemes.

• Journal of Broadcast Engineering
• /
• v.27 no.3
• /
• pp.318-331
• /
• 2022

With the recent development of hardware computing devices and software based frameworks, machine tasks using deep learning networks are expected to be utilized in various industrial fields and personal IoT devices. However, in order to overcome the limitations of high cost device for utilizing the deep learning network and that the user may not receive the results requested when only the machine task results are transmitted from the server, Collaborative Intelligence (CI) proposed the transmission of feature maps as a solution. In this paper, an efficient compression method for feature maps with vast data sizes to support the CI paradigm was analyzed and presented through experiments. This method increases redundancy by applying feature map reordering to improve compression efficiency in traditional video codecs, and proposes a feature map method that improves compression efficiency and maintains the performance of machine tasks by simultaneously utilizing image compression format and video compression format. As a result of the experiment, the proposed method shows 14.29% gain in BD-rate of BPP and mAP compared to the feature compression anchor of MPEG-VCM.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.2
• /
• pp.67-72
• /
• 2022

Recently, various intelligent application services using artificial intelligence are being actively developed. In particular, research on artificial intelligence-based real-time prediction services is being actively conducted in the manufacturing industry, and the demand for artificial intelligence services that can detect and predict fire and odors is very high. However, most of the existing detection and prediction systems do not predict the occurrence of fires and odors, but rather provide detection services after occurrence. This is because AI-based prediction service technology is not applied in existing systems. In addition, fire prediction, odor detection and odor level prediction services are services with ultra-low delay characteristics. Therefore, in order to provide ultra-low-latency prediction service, edge computing technology is combined with artificial intelligence models, so that faster inference results can be applied to the field faster than the cloud is being developed. Therefore, in this paper, we propose an LSTM algorithm-based learning model that can be used for fire prediction and odor detection/prediction, which are most required in the manufacturing industry. In addition, the proposed learning model is designed to be implemented in edge devices, and it is proposed to receive real-time sensor data from the IoT terminal and apply this data to the inference model to predict fire and odor conditions in real time. The proposed model evaluated the prediction accuracy of the learning model through three performance indicators, and the evaluation result showed an average performance of over 90%.

• Journal of the Korea Convergence Society
• /
• v.13 no.3
• /
• pp.77-89
• /
• 2022

It has been proven through many previous researches that the pretrained language model with a large corpus helps improve performance in various natural language processing tasks. However, there is a limit to building a large-capacity corpus for training in a language environment where resources are scarce. Using the Cross-lingual Post-Training (XPT) method, we analyze the method's efficiency in Korean, which is a low resource language. XPT selectively reuses the English pretrained language model parameters, which is a high resource and uses an adaptation layer to learn the relationship between the two languages. This confirmed that only a small amount of the target language dataset in the relationship extraction shows better performance than the target pretrained language model. In addition, we analyze the characteristics of each model on the Korean language model and the Korean multilingual model disclosed by domestic and foreign researchers and companies.