• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.30-33
• /
• 2023

The recent scholarly focus has been directed towards the expeditious and accurate detection of salient objects, a task that poses considerable challenges for resource-limited edge devices due to the high computational demands of existing models. To mitigate this issue, some contemporary research has favored inference speed at the expense of accuracy. In an effort to reconcile the intrinsic trade-off between accuracy and computational efficiency, we present novel model for salient object detection. Our model incorporate group-wise attentive module within the decoder of the encoder-decoder framework, with the aim of minimizing computational overhead while preserving detection accuracy. Additionally, the proposed architectural design employs attention mechanisms to generate boundary information and semantic features pertinent to the salient objects. Through various experimentation across five distinct datasets, we have empirically substantiated that our proposed models achieve performance metrics comparable to those of computationally intensive state-of-the-art models, yet with a marked reduction in computational complexity.

• Journal of Engineering Education Research
• /
• v.26 no.6
• /
• pp.40-50
• /
• 2023

This study is to examine the validity of assessing basic-level proficiency in physics among incoming engineering freshmen through item analysis and standard setting. For empirical analysis, we examined the physics subject taken by the freshman class of 2021 at K University, considering its significance for engineering students. In this study, we initially performed item analysis utilizing both classical test theory and item response theory. Subsequently, leveraging the item and test information, we employed a modified Angoff method and the Bookmark method for standard setting. Consequently, the difficulty level initially set during item development was found to be higher than the actual performance level exhibited by the students. This study highlights a discernible disparity between the expected university standard and the real proficiency level of incoming freshmen in terms of basic academic ability in physics. Based on these research findings, a comprehensive discussion on the fundamental academic competence of engineering students was conducted, underscoring the necessity for formulating a tailored learning approach leveraging the outcomes from the basic ability test.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.9
• /
• pp.2573-2589
• /
• 2023

As 5G mobile systems carry multiple services and applications, numerous user, and application types with varying quality of service requirements inside a single physical network infrastructure are the primary problem in constructing 5G networks. Radio Access Network (RAN) slicing is introduced as a way to solve these challenges. This research focuses on optimizing RAN slices within a singular physical cell for vehicle-to-everything (V2X) and enhanced mobile broadband (eMBB) UEs, highlighting the importance of adept resource management and allocation for the evolving landscape of 5G services. We put forth two unique strategies: one being offline network slicing, also referred to as standard network slicing, and the other being Online reinforcement learning (RL) network slicing. Both strategies aim to maximize network efficiency by gathering network model characteristics and augmenting radio resources for eMBB and V2X UEs. When compared to traditional network slicing, RL network slicing shows greater performance in the allocation and utilization of UE resources. These steps are taken to adapt to fluctuating traffic loads using RL strategies, with the ultimate objective of bolstering the efficiency of generic 5G services.

• ETRI Journal
• /
• v.45 no.6
• /
• pp.963-973
• /
• 2023

Non-orthogonal multiple access (NOMA) is considered a key candidate technology for next-generation wireless communication systems due to its high spectral efficiency and massive connectivity. Incorporating the concepts of multiple-input-multiple-output (MIMO) into NOMA can further improve the system efficiency, but the hardware complexity increases. This study develops an energy-efficient (EE) subchannel assignment framework for MIMO-NOMA systems under the quality-of-service and interference constraints. This framework handles an energy-efficient co-training-based semi-supervised learning (EE-CSL) algorithm, which utilizes a small portion of existing labeled data generated by numerical iterative algorithms for training. To improve the learning performance of the proposed EE-CSL, initial assignment is performed by a many-to-one matching (MOM) algorithm. The MOM algorithm helps achieve a low complex solution. Simulation results illustrate that a lower computational complexity of the EE-CSL algorithm helps significantly minimize the energy consumption in a network. Furthermore, the sum rate of NOMA outperforms conventional orthogonal multiple access.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.47 no.1
• /
• pp.1-8
• /
• 2024

This study analyzed the impact of Higher-order resources on profit sustainability for domestic companies using a mathematical statistical model. Higher-order resources refer to resources that do not directly affect profits but influence other resources that directly contribute to profits. As a result of analysis using 30 years of actual data from more than 650 domestic companies, the average duration of competitive advantage including high-order resources was found to be about twice as long as the period suggested by the autoregressive model excluding high-order resources. Through this, if companies want to earn more profits over a long period of time than their competitors, they must not only possess resources that are more valuable, rare, difficult to imitate, and non-substitutable compared to their competitors, but also that higher-order resources can contribute to changes in these resources over time. It was confirmed that it must lead the long-term profit difference. High-level resources include strategic planning, mergers and acquisitions (M&A) capabilities, and good forecasting.

• Applied Chemistry for Engineering
• /
• v.35 no.4
• /
• pp.341-345
• /
• 2024

Common adhesives contain harmful chemicals, posing risks to humans and the environment. Basil seeds, an underutilized resource, can offer a solution. We extracted mucilage from basil seeds and found that altering the basil seed mucilage (BSM) to water ratio allowed us to control the adhesive strength. As the concentration of BSM increased and the water evaporated, adhesive strength improved. BSM can serve as both an adhesive and pressure-sensitive adhesive. This research showcases BSM's impressive strength and stability, making it a promising eco-friendly industrial adhesive option from plant-based sources, revolutionizing the adhesive industry.

• International Journal of Computer Science & Network Security
• /
• v.24 no.9
• /
• pp.150-156
• /
• 2024

With the advent of personalized search engines, a myriad of approaches came into practice. With social media emergence the personalization was extended to different level. The main reason for this preference of personalized engine over traditional search was need of accurate and precise results. Due to paucity of time and patience users didn't want to surf several pages to find the result that suits them most. Personalized search engines could solve this problem effectively by understanding user through profiles and histories and thus diminishing uncertainty and ambiguity. But since several layers of personalization were added to basic search, the response time and resource requirement (for profile storage) increased manifold. So it's time to focus on optimizing the layered architectures of personalization. The paper presents a layout of the multi agent based personalized search engine that works on histories and profiles. Further to store the huge amount of data, distributed database is used at its core, so high availability, scaling, and geographic distribution are built in and easy to use. Initially results are retrieved using traditional search engine, after applying layer of personalization the results are provided to user. MongoDB is used to store profiles in flexible form thus improving the performance of the engine. Further Weighted Sum model is used to rank the pages in personalization layer.

• Journal of Korea Technology Innovation Society
• /
• v.21 no.2
• /
• pp.587-614
• /
• 2018

The Korean government is aiming to strengthen industrial and national competitiveness through the promotion of cooperation between universities, public research institutes and industry and vitalization of technology transfer. In 2013 and 2014, the Ministry of Trade, Industry and Energy and Ministry of Science, ICT and Future Planning have announced policies to support SMEs by public research organizations. In addition, in 2015, the 'Korean Fraunhofer support system', which pay government subsidies according to the amounts of private R&D funds was adopted. However, there are some concern about the government's policies. There is yet disclosed how these policies affect technology transfer because industrial R&D funding has not been activated in Korea unlike German. Therefore this paper analyzes effects of relationship between universities, public research institutes and external organizations on performance of technology transfer based on the Triple Helix Model. Empirical results show that the relationship with the government has a significant impact on the resource security and the relationship with the industry has a significant effect on the diffusion of the performance. In addition, a public research institute was selected and case analysis was conducted to suggest policy implications for improving the technology transfer performance of universities and public research institutes.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.46 no.4
• /
• pp.95-103
• /
• 2009

In this paper, a generation method of transmit and receive beamforming vectors based on base station cooperation is proposed which maximizes the user SINR in mobile cellular multi-user MIMO systems. There are two main sources of interference which deteriorate the performance of the system, i.e. the inter-user interference caused by the usage of the same radio resource for multiple users in the system, and the inter-cluster interference from neighboring base stations which are not participating in cooperative transmission. The proposed scheme cancels out the inter-user interference by using the block diagonalization(BD) method, and mitigate the inter-cluster interference by using optimal transmit and receive beamforming vectors based on optimal combining(OC) with the statistic information of inter-cluster interference. We perform computer simulations to verify the performance of the proposed scheme, and compare the result to the conventional performance obtained from utilizing the receiver side information only or utilizing the information from neither sides. The performance evaluations are conducted not only over the independent MIMO channels, but over correlated MIMO channels to demonstrate the robustness of the proposed scheme over the channels with correlation among antennas.

• Journal of Adhesion and Interface
• /
• v.22 no.1
• /
• pp.1-7
• /
• 2021

The importance of high capacity energy storage devices has recently emerged for stable power supply through renewable energy generation. From this point of view, the Na-air battery (NAB), which is a next-generation secondary battery, is receiving huge attention because it can realize a high capacity through abundant and inexpensive raw materials. In this study, activated carbon-based catalysts for hybrid type Na-air batteries were prepared and their characteristics were compared and analysed. In particular, from the viewpoint of resource recycling, activated carbon (Orange-C) was prepared using discarded orange peel, and performance was compared with Vulcan carbon, which is widely used. In addition, a Pt/C catalyst (homemade-Pt/C, HM-Pt/C) was synthesized using a modified polyol method to check whether the prepared activated carbon can be used as a supported catalyst, and a commercial Pt/C catalyst (Commercial Pt/C) and electrochemical performance were compared. The prepared Orange-C exhibited a typical H3 type BET isotherm, which is evidence that micropore and mesopore exist. In addition, in the case of HM-Pt/C, it was confirmed through TEM analysis that Pt particles were evenly distributed on the activated carbon supported catalyst. In particular, the HM-Pt/C-based NAB showed the smallest voltage gap (0.224V) and good voltage efficiency (92.34%) in the 1st galvanostatic charge-discharge test. In addition, the cycle performance test conducted for 20 cycles showed the most stable performance.