• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.6
• /
• pp.850-858
• /
• 2022

To deploy Gate Recurrent Units (GRU) on resource-constrained embedded devices, this paper presents a reconfigurable FPGA-based GRU accelerator that enables structured compression. Firstly, a dense GRU model is significantly reduced in size by hybrid quantization and structured top-k pruning. Secondly, the energy consumption on external memory access is greatly reduced by the proposed reuse computing pattern. Finally, the accelerator can handle a structured sparse model that benefits from the algorithm-hardware co-design workflows. Moreover, inference tasks can be flexibly performed using all functional dimensions, sequence length, and number of layers. Implemented on the Intel DE1-SoC FPGA, the proposed accelerator achieves 45.01 GOPs in a structured sparse GRU network without batching. Compared to the implementation of CPU and GPU, low-cost FPGA accelerator achieves 57 and 30x improvements in latency, 300 and 23.44x improvements in energy efficiency, respectively. Thus, the proposed accelerator is utilized as an early study of real-time embedded applications, demonstrating the potential for further development in the future.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.15 no.2
• /
• pp.124-133
• /
• 2022

With the development of hardware technology and the advancement of numerical model methods, more precise weather forecasts can be carried out. In this paper, we propose a Unicon Optimization scheme combining Loop Vectorization, Dependency Vectorization, and Code Modernization to optimize and increase Maintainability the Unicon source contained in SCAM, a simplified version of CESM, and present an overall SCAM structure. This paper tested the unicorn optimization scheme in the SCAM structure, and compared to the existing source code, the loop vectorization resulted in a performance improvement of 3.086% and the dependency vectorization of 0.4572%. And in the case of Unicorn Optimization, which applied all of these, the performance improvement was 3.457% compared to the existing source code. This proves that the Unicorn Optimization technique proposed in this paper provides excellent performance.

• Journal of Korea Society of Industrial Information Systems
• /
• v.15 no.2
• /
• pp.1-9
• /
• 2010

Feistel and SPN are the two main structures in a block cipher. Feistel is a symmetric structure which has the same structure in encryption and decryption, but SPN is not a symmetric structure. In this paper, we propose a SPN which has a symmetric structure in encryption and decryption. The whole operations of proposed algorithm are composed of the even numbers of N rounds where the first half of them, 1 to N/2 round, applies a right function and the last half of them, (N+1)/2 to N round, employs an inverse function. And a symmetry layer is located in between the right function layer and the inverse function layer. In this paper, AES encryption and decryption function are selected for the right function and the inverse function, respectively. The symmetric layer is composed with simple matrix and round key addition. Due to the simplicity of the symmetric SPN structure in hardware implementation, the proposed modified AES is believed to construct a safe and efficient cipher in Smart Card and RFID environments where electronic chips are built in.

• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.3
• /
• pp.129-136
• /
• 2022

The most important part of the smart livestock building system is to maintain a breeding environment so that livestock can grow to high quality despite changes in the internal and external atmospheric environment. Especially, it is very important to maintain the temperature and humidity in the livestock building because various diseases occur during the summer and winter. To manage the environment suitable for livestock, a smartfarm system for livestock building is applied, but it is very expensive. In this study, we propose a hardware design and control method for low cost system based on HMI and fuzzy control. To evaluate the performance of the proposed system, we did a simulation experiment in the atmospheric conditions of summer and winter. As a result, it showed the performance of minimizing the temperature and humidity stress of livestock. And when applied to the livestock building, the proposed system showed stable control performance even in the change of the external atmospheric environment. Therefore, as with these results, if proposed system in this study is applied to the smart farm system, it will be effective in managing the environment of livestock building.

• Journal of Digital Convergence
• /
• v.20 no.1
• /
• pp.185-190
• /
• 2022

Bluetooth is a common wireless technology that is widely used as a connection medium between various consumer electronic devices. The Bluetooth receiver usually adopts a Viterbi algorithm to improve signal-to-noise ratio performance, but requires complex hardware and calculations for continuous search and estimation for the irrational modulation indexes at the transmission. This paper proposes a non-coherent maximum estimation based 8-State Viterbi FSM to solve these complexity problems. The proposed optimal Viterbi FSM can detect Gaussian frequency-shfit keying symbol without any prior information and estimation for the modulation indexes. The HV1/HV2 packets are used for the estimation of the proposed algorithm and the simulation results have shown performance improvements with about 2dB for 10^-3 BER compared to other ideal approaches such as decision direct method.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.7 no.6
• /
• pp.871-880
• /
• 2017

Recently, service based on internet is inter-connected and integrated with a variety of connection. This kind of internet of things consist of heterogenous devices such as sensor node, devices and end-to end equipment which used in conventional protocols and services. The representative system is healthcare system. From healthcare appliance used by IoT, patient and doctor can utilize healthcare information with safety and high speed management. It is very convenient management to operate mobility. But it induced security and vulnerability issues because it has small memory capacity, low power supply and low computing power. This made impossible to implement security algorithm with embedded engine based on hardware. Nowdays, we can't realize conventional standard algorithm due to these kinds of reasons. From the critical issues, it occurred security and vulnerability issues. Therefore, we analysed and compared with conventional method and proposed techniques. Finally, we evaluated security issues and requirement for end-to-end point healthcare system based on internet of things.

• Journal of Information Technology Applications and Management
• /
• v.29 no.1
• /
• pp.47-57
• /
• 2022

The purpose of this study is to analyze the current status of augmented reality picture books, which have been steadily developed since 2010, as a genre of electronic picture books, and to reveal how children's immersion in augmented reality picture books differs from paper picture books. To this end, 30 augmented reality picture books on the market were analyzed according to genre, life theme, implementation method, and augmented reality scene ratio. As a result of the study, it was found that the genre of picture books was in the order of information fairy tales, daily fairy tales, and historical fairy tales, and there were no traditional or fantasy fairy tales. Animals and plants accounted for about half of the life topics, and in some cases, there were only a few or no other life topics. In the augmented reality implementation scene, it consisted of only one page in the early days, but all pages are now possible to implement augmented reality due to technology development, production cost reduction, and improved hardware performance of smartphones. It was found that the augmented reality implementation method used CD-ROM in the early days, but gradually became possible to implement using only mobile phones and tablets that were easy for readers to access. In addition, after presenting four picture books to eight 5-year-old infants, the immersion time was measured and the immersion behavior was observed. As a result, augmented reality picture books showed higher immersion[immersion time, immersion behavior] than paper picture books, but compared by literature genre, life fairy tales were higher in paper picture books and natural fairy tales in augmented reality picture books. It was higher when presenting augmented reality picture books after presenting paper picture books according to the order of presentation of picture book types. The results of this study suggest that more diverse life topics and augmented reality picture books in the genre of children's literature should be developed to increase the utilization of augmented reality picture books. In addition, considering that there are differences in immersion between types, literary genres, and reading experience [presentation order], it is expected to increase the educational effect by using picture books complementarily.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.16 no.5
• /
• pp.307-314
• /
• 2023

As supercomputing and hardware technology advances, climate prediction models are improving. The Korean Meteorological Administration adopted GloSea5 from the UK Met Office and now operates an updated GloSea6 tailored to Korean weather. Universities and research institutions use Low-GloSea6 on smaller servers, improving accessibility and research efficiency. In this paper, profiling Low-GloSea6 on smaller servers identified the tri_sor_dp_dp subroutine in the tri_sor.F90 atmospheric model as a CPU-intensive hotspot. Applying linear regression, a type of machine learning, to this function showed promise. After removing outliers, the linear regression model achieved an RMSE of 2.7665e-08 and an MAE of 1.4958e-08, outperforming Lasso and ElasticNet regression methods. This suggests the potential for machine learning in optimizing identified hotspots during Low-GloSea6 execution.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.17 no.1
• /
• pp.38-46
• /
• 2024

In general, EEG signal analysis has been the subject of several studies due to its ability to provide an objective mode of recording brain stimulation, which is widely used in brain-computer interface research with applications in medical diagnosis and rehabilitation engineering. In this study, we developed EEG reception hardware to measure electroencephalograms and implemented a processing system, classifying it into server and data processing. It was conducted as an intermediate-stage research on the implementation of a brain-computer interface using electroencephalograms, and was implemented in the form of predicting the user's arm movements according to measured electroencephalogram data. Electroencephalogram measurements were performed using input from four electrodes through an analog-to-digital converter. After sending this to the server through a communication process, we designed and implemented a system flow in which the server classifies the electroencephalogram input using a convolutional neural network model and displays the results on the user terminal.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.4
• /
• pp.589-594
• /
• 2024

In this paper, we introduce the results of the Sounding Rocket LTE communication test using the LTE-Cat.M1 module. The developed LTE data transmission/reception system consists of Mission-Mounted Equipment(Payload) and Ground Observation Equipment(GOE), and the delay rate was secured based on the time between data measured when received from the GOE by constantly transmitting data from the Payload at a speed of 10 Hz. In order to increase the accuracy of the actual flight test, ground network delay rate tests, hardware internal delay rate tests, and ground tests were performed. As a result of the flight test, it was confirmed that the handover failed in the upward phase and the communication was lost for 13 seconds, and then the parachute was deployed and the communication was reconnected in a situation with a constant positional displacement. LTE-Cat.M1 technology is expected to be utilized for descent phase observation missions or data backup during Sounding Rocket missions.