• ETRI Journal
• /
• 제37권6호
• /
• pp.1087-1095
• /
• 2015

Mobility management is important in mobile cellular networks. In this study, we considered an enhanced location-based registration (ELR) method. In the ELR method, even when a mobile phone enters a cell to find that the cell is already on its list (of visited cells) and then updates its main counter, it does not remove any cells from the list (memory space permitting), which gives better performance than the location-based registration (LR) method. However, the location registration cost of the ELR method is still high, and there is a lot of room for improvement with regards to this matter. We now propose an improved version of the ELR method; namely, the improved ELR (iELR). In the iELR method, when a mobile phone enters a cell to find that the cell counter value is less than the main counter value, or when a mobile phone enters a cell to register its location, it updates the main counter and the cell counter values as much as possible to reduce the future need for registrations. We show that our proposed iELR method provides better performance than the ELR method.

• 전기학회논문지P
• /
• 제57권2호
• /
• pp.135-139
• /
• 2008

MFISFETs with platinum electrode on the $LiNbO_3$/aluminum nitride/Si(100) structures were successfully fabricated and the properties of the FETs have been discussed. $I_D-V_G$ characteristics of MFISFETs for linear region (that is, 0.1 V of the drain voltage) showed hysteresis loop with a counter-clockwise trace due to the ferroelectric nature of $LiNbO_3$ films. A memory window (i.e., threshold voltage shift) of the fabricated device was about 2[V] for a sweep from -4 to +4[V]. The estimated field-effect electron mobility and transconductance on a linear region were 530[$cm^2/V{\cdot}s$] and 0.16[mS/mm], respectively. The drain current of 27[${\mu}A$] on the "on" state was more than 3 orders of magnitude larger than that of 30[nA] on the "off" state at the same "read" gate voltage of l.5[V], which means the memory operation of the MFISFET.

• 해양환경안전학회지
• /
• 제29권5호
• /
• pp.505-511
• /
• 2023

본 논문에서는 소형어선의 운동 응답을 예측하기 위해 딥러닝 모델을 구축하였다. 크기가 다른 두 소형어선을 대상으로 유체동역학 성능을 평가하여 데이터세트를 확보하였다. 딥러닝 모델은 순환 신경망 기법의 하나인 장단기 메모리 기법(LSTM, Long Short-Term Memory)을 사용하였다. 딥러닝 모델의 입력 데이터는 6 자유도 운동 및 파고의 시계열 데이터를 사용하였으며, 출력 라벨로는 6 자유도 운동의 시계열 데이터로 선정하였다. 최적 LSTM 모델 구축을 위해 hyperparameter 및 입력창 길이의 영향을 평가하였다. 구축된 LSTM 모델을 통해 입사파 방향에 따른 시계열 운동 응답을 예측하였다. 예측된 시계열 운동 응답은 해석 결과와 전반적으로 잘 일치함을 확인할 수 있었다. 시계열의 길이가 길어짐에 따라서 예측값과 해석 결과의 차이가 발생하는데, 이는 장기 데이터에 따른 훈련 영향도가 감소 됨에 따라 나타난 것으로 확인할 수 있다. 전체 예측 데이터의 오차는 약 85% 이상의 데이터가 10% 이내의 오차를 보였으며, 소형어선의 시계열 운동 응답을 잘 예측함을 확인하였다. 구축된 LSTM 모델은 소형어선의 모니터링 및 경보 시스템에 활용될 수 있을 것으로 기대한다.

• 한국전기전자재료학회논문지
• /
• 제25권2호
• /
• pp.85-90
• /
• 2012

In this paper, we compared and analyzed 3D silicon-oxide-nitride-oxide-silicon (SONOS) multi layer flash memory devices fabricated on nitride or oxide layer, respectively. The device fabricated on nitride layer has inferior electrical properties than that fabricated on oxide layer. However, the device on nitride layer has faster program / erase speed (P/E speed) than that on the oxide layer, although having inferior electrical performance. Afterwards, to find out the reason why the device on nitride has faster P/E speed, 1/f noise analysis of both devices is investigated. From gate bias dependance, both devices follow the mobility fluctuation model which results from the lattice scattering and defects in the channel layer. In addition, the device on nitride with better memory characteristics has higher normalized drain current noise power spectral density ($S_{ID}/I^2_D$>), which means that it has more traps and defects in the channel layer. The apparent hooge's noise parameter (${\alpha}_{app}$) to represent the grain boundary trap density and the height of grain boundary potential barrier is considered. The device on nitride has higher ${\alpha}_{app}$ values, which can be explained due to more grain boundary traps. Therefore, the reason why the devices on nitride and oxide have a different P/E speed can be explained due to the trapping/de-trapping of free carriers into more grain boundary trap sites in channel layer.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제18권2호
• /
• pp.456-477
• /
• 2024

With information technology's rapid development, the Internet faces serious security problems. Studies have shown that malware has become a primary means of attacking the Internet. Therefore, adversarial samples have become a vital breakthrough point for studying malware. By studying adversarial samples, we can gain insights into the behavior and characteristics of malware, evaluate the performance of existing detectors in the face of deceptive samples, and help to discover vulnerabilities and improve detection methods for better performance. However, existing adversarial sample generation methods still need help regarding escape effectiveness and mobility. For instance, researchers have attempted to incorporate perturbation methods like Fast Gradient Sign Method (FGSM), Projected Gradient Descent (PGD), and others into adversarial samples to obfuscate detectors. However, these methods are only effective in specific environments and yield limited evasion effectiveness. To solve the above problems, this paper proposes a malware adversarial sample generation method (PixGAN) based on the pixel attention mechanism, which aims to improve adversarial samples' escape effect and mobility. The method transforms malware into grey-scale images and introduces the pixel attention mechanism in the Deep Convolution Generative Adversarial Networks (DCGAN) model to weigh the critical pixels in the grey-scale map, which improves the modeling ability of the generator and discriminator, thus enhancing the escape effect and mobility of the adversarial samples. The escape rate (ASR) is used as an evaluation index of the quality of the adversarial samples. The experimental results show that the adversarial samples generated by PixGAN achieve escape rates of 97%, 94%, 35%, 39%, and 43% on the Random Forest (RF), Support Vector Machine (SVM), Convolutional Neural Network (CNN), Convolutional Neural Network and Recurrent Neural Network (CNN_RNN), and Convolutional Neural Network and Long Short Term Memory (CNN_LSTM) algorithmic detectors, respectively.

• 한국재료학회지
• /
• 제21권4호
• /
• pp.192-195
• /
• 2011

The electro-deposition of compound semiconductors has been attracting more attention because of its ability to rapidly deposit nanostructured materials and thin films with controlled morphology, dimensions, and crystallinity in a costeffective manner (1). In particular, low band-gap $A_2B_3$-type chalcogenides, such as $Sb_2Te_3$ and $Bi_2Te_3$, have been extensively studied because of their potential applications in thermoelectric power generator and cooler and phase change memory. Thermoelectric $Sb_xTe_y$ films were potentiostatically electrodeposited in aqueous nitric acid electrolyte solutions containing different ratios of $TeO_2$ to $Sb_2O_3$. The stoichiometric $Sb_xTe_y$ films were obtained at an applied voltage of -0.15V vs. SCE using a solution consisting of 2.4 mM $TeO_2$, 0.8 mM $Sb_2O_3$, 33 mM tartaric acid, and 1M $HNO_3$. The stoichiometric $Sb_xTe_y$ films had the rhombohedral structure with a preferred orientation along the [015] direction. The films featured hole concentration and mobility of $5.8{\times}10^{18}/cm^3$ and $54.8\;cm^2/V{\cdot}s$, respectively. More negative applied potential yielded more Sb content in the deposited $Sb_xTe_y$ films. In addition, the hole concentration and mobility decreased with more negative deposition potential and finally showed insulating property, possibly due to more defect formation. The Seebeck coefficient of as-deposited $Sb_2Te_3$ thin film deposited at -0.15V vs. SCE at room temperature was approximately 118 ${\mu}V/K$ at room temperature, which is similar to bulk counterparts.

• 전기전자학회논문지
• /
• 제27권3호
• /
• pp.265-272
• /
• 2023

본 논문에서는 FMCW(Frequency Modulated Continuous Wave) SAR(Synthetic Aperture Radar) 기반 실시간 영상 형성을 위해 RDA(Range Doppler Algorithm)의 GPU 가속 커널을 개발하였다. Host와 GPU device 사이의 데이터 전송 시간을 최소화하기 위해 pinned 메모리를 사용하였고, 데이터의 전송 횟수를 최소화하기 위해 모든 RDA 연산을 GPU에서 수행하도록 커널을 구성하였다. FMCW 드론 SAR 실험을 통해 데이터셋를 획득하였고, intel i7-9700K CPU, 32GB RAM과 Nvidia RTX 3090 GPU 환경에서 GPU의 가속 효과를 측정하였다. Host-device간 데이터 전송 시간을 포함했을 경우 CPU 대비 최대 3.41배 가속된 것으로 측정되었고, 데이터 전송 시간을 포함하지 않고 연산의 가속 효과만을 측정했을 때, 최대 156배 가속 가능함을 확인할 수 있었다.

• 대한전자공학회논문지TC
• /
• 제43권4호
• /
• pp.1-12
• /
• 2006

본 논문에서는 IP기반의 통합망에서의 효율적인 모바일 멀티미디어 통신 프로토콜, 단말 소프트웨어 플랫폼 및 모바일 VoIP를 제안한다. 제안하는 모바일 멀티미디어 통신 프로토콜은 터널분할기반 모바일 자원예약 프로토콜이라 부르며, 이는 고속 이동성 지원을 위해서 터널분할 방식의 Mobile IP와 RSVP를 연동하고 있다. 또한, 통신품질을 유지하면서 끊어짐없는 핸드오버를 단말 플랫폼에서도 지원해야 하기 때문에 모바일 통신품질지원 모듈들을 단말플랫폼용으로 개발하고 공유메모리 방식을 통해 상호연동시킨다. 실험망으로는 제안하는 프로토콜을 실장한 코어네트워크와 무선랜 기반의 엑세스 네트워크들로 구성한다. 자체 개발한 모바일 VoIP를 이용하여 제안하는 기술들의 기능성과 성능을 다양한 이동성 실험을 통해 검증한다. 결론적으로 제안하는 방식이 기존 표준적 방식에 비해 30% 이하의 자원예약기반 핸드오프지연시간을 지원하여 CDMA 휴대전화와 동일한 수준의 음성통화품질을 제공한다.

• Transactions on Electrical and Electronic Materials
• /
• 제14권5호
• /
• pp.250-253
• /
• 2013

In this paper, we fabricated 3D pillar type silicon-oxide-nitride-oxide-silicon (SONOS) devices for high density flash applications. To solve the limitation between erase speed and data retention of the conventional SONOS devices, bandgap-engineered (BE) tunneling oxide of oxide-nitride-oxide configuration is integrated with the 3D structure. In addition, the tunneling oxide is modulated by another method of $N_2$ ion implantation ($N_2$ I/I). The measured data shows that the BE-SONOS device has better electrical characteristics, such as a lower threshold voltage ($V_{\tau}$) of 0.13 V, and a higher $g_{m.max}$ of 18.6 ${\mu}A/V$ and mobility of 27.02 $cm^2/Vs$ than the conventional and $N_2$ I/I SONOS devices. Memory characteristics show that the modulated tunneling oxide devices have fast erase speed. Among the devices, the BE-SONOS device has faster program/erase (P/E) speed, and more stable endurance characteristics, than conventional and $N_2$ I/I devices. From the flicker noise analysis, however, the BE-SONOS device seems to have more interface traps between the tunneling oxide and silicon substrate, which should be considered in designing the process conditions. Finally, 3D structures, such as the pillar type BE-SONOS device, are more suitable for next generation memory devices than other modulated tunneling oxide devices.

• 한국정밀공학회지
• /
• 제30권1호
• /
• pp.47-52
• /
• 2013

In order to operate a search and rescue robot in hazardous area, the robot requires high mobility and adaptable locomotion for moving in unpredictable environments. In this paper, we propose the deformable soft wheel robot that can produce three kinds of driving modes; caterpillar driving mode, normal wheel driving mode, legged-wheel driving mode. The robot changes its driving mode as it faces the various obstacles such as a small gap, stairs etc. Soft film and composite materials are used for fabrication of deformable wheel structure and Shape Memory Alloy (SMA) coil spring actuators are attached on the structure as an artificial muscle. Film lamination and an composite manufacturing process is introduced and the robot design is required to be modified and compromised to applying the manufacturing process. The prototype is developed and tested for verifying feasibility of the deformable wheel locomotion.