• ALGAE
• /
• v.33 no.1
• /
• pp.127-141
• /
• 2018

Low efficiency in microalgal biomass production was largely attributed to the low density of algal cell cultures. Though mutations that reduced the level of chlorophyll or pigment content increased efficiency of photon usage and thus the cell-culture density under high-illumination growth conditions (e.g., >$500{\mu}mol\;photon\;m^{-2}\;s^{-1}$), it was unclear whether algae could increase cell-culture density under low-illumination conditions (e.g., ${\sim}50{\mu}mol\;photon\;m^{-2}\;s^{-1}$). To address this question, we performed forward genetic screening in Chlamydomonas reinhardtii. A pool of >1,000 insertional mutants was constructed and subjected to continuous subcultures in shaking flasks under low-illumination conditions. Complexity of restriction fragment length polymorphism (RFLP) pattern in cultures indicated the degree of heterogeneity of mutant populations. We showed that the levels of RFLP complexity decreased when cycles of subculture increased, suggesting that cultures were gradually populated by high cell-culture density (HCD) strains. Analysis of the 3 isolated HCD mutants after 30 cycles of subcultures confirmed that their maximal biomass production was 50-100% higher than that of wild type under low-illumination. Furthermore, levels of chlorophyll content in HCD mutant strains were similar to that of wild type. Inverse polymerase chain reaction analysis identified the locus of insertion in two of three HCD strains. Molecular and transcriptomic analyses suggested that two HCD mutants were a result of the gain-of-function phenotype, both linking to the abnormality of mitochondrial functions. Taken together, our results demonstrate that HCD strains can be obtained through continuous subcultures under low illumination conditions.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.11 no.6
• /
• pp.638-644
• /
• 2018

Relay systems have recently gained attentions because of its capability of cell coverage extension and the power gain as the one of key technologies for 5G. Relays can be exploited for small-cell base stations and the autonomous network, where communication devices communicate with each other cooperatively. Therefore, the relay technology is expected to enable the low power and large capacity communication. In order to maximize the benefits of using a limited number of relays, the efficient relay selection method is required. Especially, when two nodes exchange their data with each other via relay, the relay selection can maximize the average data rate by the spatial location of the relay. For this purpose, the average data rate is analyzed first according to the relay selection. In this paper, we analyzed the average data rate when two nodes exchange their data via dual-hop decode and forward relaying considering the interference by the concurrent transmission under Nakagami-m fading channel. The correctness of the analysis is verified by the Monte Carlo simulation. The results show that the concurrent transmission is superior to the non-concurrent transmission in the high required data rate region rather than in the low required data rate region.

• International Journal of Fluid Machinery and Systems
• /
• v.8 no.4
• /
• pp.264-273
• /
• 2015

This presentation describes an experimental approach for the detection of cavitation in hydraulic machines by use of ultrasonic signal analysis. Instead of using the high frequency pulses (typically 1MHz) only for transit time measurement different other signal characteristics are extracted from the individual signals and its correlation function with reference signals in order to gain knowledge of the water conditions. As the pulse repetition rate is high (typically 100Hz), statistical parameters can be extracted of the signals. The idea is to find patterns in the parameters by a classifier that can distinguish between the different water states. This classification scheme has been applied to different cavitation sections: a sphere in a water flow in circular tube at the HSLU in Lucerne, a NACA profile in a cavitation tunnel and two Francis model test turbines all at LMH in Lausanne. From the signal raw data several statistical parameters in the time and frequency domain as well as from the correlation function with reference signals have been determined. As classifiers two methods were used: neural feed forward networks and decision trees. For both classification methods realizations with lowest complexity as possible are of special interest. It is shown that two to three signal characteristics, two from the signal itself and one from the correlation function are in many cases sufficient for the detection capability. The final goal is to combine these results with operating point, vibration, acoustic emission and dynamic pressure information such that a distinction between dangerous and not dangerous cavitation is possible.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.6A
• /
• pp.420-425
• /
• 2012

A forward error correction (FEC) coding techniques is one of time diversity techniques with which the effect of channel impairments due to noise and fading are spreaded over independently, and thus the performance could be improved. Therefore, the performance of the FEC scheme can be maximized if we minimize the correlation of channel information across over a codeword. In this paper, we propose a block turbo code with the maximized time diversity effect which may be reduced due to utilization of high order modulation schemes and due to transmission over a comparatively fast fading environment. Especially, we propose a very simple formula to calculate the address of coded bit allocation, and thus we do not need any additional outer interleavers, i.e., inter-codeword interleavers. The simulation resuts investigated in this paper reveal that the proposed scheme can provide the performance gain of more than a few decibels compared to the conventional schemes.

• Korean Journal of Applied Biomechanics
• /
• v.16 no.1
• /
• pp.55-63
• /
• 2006

The purpose of this study is to prove the kinematical characteristics of Deff motion, the high bar performance, in terms of flying phases so that we can provide basic sources for improving gymnastic performance. To do this, we selected and analyzed the performance of two athletes who did Deff motion in the high bar competition of male artistic gymnastic in the 22nd Universiade 2003 Daegu. We drew the conclusions from the kinematical factors that were came out through analyzing three-dimensional cinematography of the athletes' movements, by using a high speed video camera. To make a successful performance, a performer releases the bar at a height of a high bar vertically and at a height of 82cm horizontally, and the flying performance should be made without moving forward, as maintaining the proper balance, in order to rise over 118cm high during the flying phase. When the performer is releasing the bar, an increase of the vertical speed in the center of the body and extension of a knee joint and a hip joint contribute to increasing a flying height. And when the moving body is twisted, leaning to left side is caused by the winding movement of a knee joint, which causes an unstable bar grasp. To grasp the bar stably, just before releasing the performer should gain propulsive force from twisting rotation through increasing the speed of shoulder rotation. And before the peak point, the performer should make sure of a body rotation distance over $164^{\circ}$ so that he or she can do an aerial rotary performance smoothly. When grasping the high bar, the center of the body should be above the bar and the angle of shoulder rotation should be maintained close to $540^{\circ}$ simultaneously. he high point performance(S1) has more speed on an ascending phase and less speed on a descending phase than the low point performance (S2). At the peak point, both the rotation angle of the body and that of the shoulder in high point performance are big as well. In conclusion, it is shown that a performer can make a jump toward the high bar easily with the body straight because the performer can hold the upper part of the body erect early in a descending phase.

• Korean Journal of Materials Research
• /
• v.12 no.6
• /
• pp.470-475
• /
• 2002

The effects of the conditions of the collector formation on electrical characteristics of the packaged SiGe hetero-junction bipolar transistors (HBT) were investigated. While the DC characteristics of SiGe HBTs such as IV characteristic, forward current gain, Early voltage, and breakdown voltage were hardly changed after packaging, the AC characteristics such as $f_{\tau}\; and\; f_{max}$ were degraded severely. With the rise of the collector concentration, the break-down voltage decreased but the $f_{\tau}$ increased. Additionally, $\beta$ and $f_{\tau}$ values were kept high in the range of elevated collector current due to the increase of the critical current density for the onset of the Kirk effect. The devices As implanted before the collector deposition showed lower breakdown voltage and higher $f_{\tau}$ than the others, which seems to be originated from the As up-diffusion resulting in the thinner collector.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.10
• /
• pp.4662-4677
• /
• 2018

The dynamic opportunistic device-to-device (DO-D2D) network will frequently emerge in the fifth generation (5G) wireless communication due to high-density and fast-moving mobile devices. In order to improve the Quality of Experience (QoE) of users with different computing capacity devices in the DO-D2D network, in this paper, we focus on the study of how to reduce the packets retransmission delay and satisfy heterogeneous devices. To select as many devices as possible to transmit simultaneously without interference, the concurrent transmitters-selecting algorithm is firstly put forward. It jointly considers the number of packets successfully received by each device and the device's connectivity. Then, to satisfy different devices' demands while primarily ensuring the base-layer packets successfully received by all the devices, the layer-cooperation instantly decodable network coding is presented, which is used to select transmission packets combination for each transmitter. Simulation results illustrate that there is an appreciable retransmission delay gain especially in the poor channel quality network compared to the traditional base-station (BS) retransmission algorithm. In addition, our proposed algorithms perform well to satisfy the different demands of users with heterogeneous devices.

• Clinics in Shoulder and Elbow
• /
• v.17 no.2
• /
• pp.68-76
• /
• 2014

Background: Recently, reverse total shoulder arthroplasty (RTSA) has been accepted as a main treatment option in irreparable massive rotator cuff tear with cuff arthropathy. The purpose of this study was to evaluate the early complication incidence and the preliminary clinical results of RTSAs performed in single institute. Methods: Fifty-seven RTSAs (56 patients) were performed between April 2011 and March 2013. The indications for RTSA were cuff tear arthropathy and irreparable massive rotator cuff tear with or without pseudoparalysis. Exclusion criteria were revision, preoperative infections and fractures. At final follow-up, 45 shoulders were enrolled. Mean follow-up duration was 12.5 months (range, 6-27 months). The mean age at the time of surgery was 73.6 years (range, 58-87 years). All the patients were functionally accessed via Constant score, American Shoulder and Elbow Surgeons (ASES) score, pain and functional visual analogue scale (VAS) scores and active range of motion. Complications were documented as major and minor. Major complications include fractures, infections, dislocations, nerve palsies, aseptic loosening of humeral or glenoid components, or glenoid screw problems. Minor complications include radiographic scapular notching, hematomas, heterotopic ossification, algodystrophy, intraoperative dislocations, intraoperative cement extravasation, or radiographic lucent lines of the glenoid. Results: The mean Constant score increased from 31.4 to 53.8 (p < 0.001). The pain and functional VAS scores improved (5.2 to 2.7, p < 0.001, 4.0 to 6.7, p < 0.001) and active forward flexion improved from $96.9^{\circ}$ to $125.6^{\circ}$ (p = 0.011). One or more complications occurred in 16 (35.6%) of 45 shoulders, with one failure (2.2%) resulting in the removal of implants by late infection. The single most common complication was scapular notching (9 [20%]). There were 4 (8.9%) axillary nerve palsies postoperatively (n=3: transient n. palsy, n=1: Symptom existed at 11 months postoperatively but improving). Conclusions: In a sort term follow-up, RTSA provided substantial gain in overall function. Most common early complications were scapular notching and postoperative neuropathy. Although overall early complication rate was as high as reported by several authors, most of the complications can be observable without compromise to patients' clinical outcome. Long term follow-up is required to clarify the clinical result and overall complication rate.

• Korean Journal of Optics and Photonics
• /
• v.27 no.1
• /
• pp.1-17
• /
• 2016

Over the past two decades, fiber-based lasers have made remarkable progress, now having reached power levels exceeding kilowatts and drawing a huge amount of attention from academy and industry as a replacement technology for bulk lasers. In this paper we review the significant factors that have led to the progress of fiber lasers, such as gain-fiber regimes based on ytterbium-doped silica, optical pumping schemes through the combination of laser diodes and double-clad fiber geometries, and tandem schemes for minimizing quantum defects. Furthermore, we discuss various power-limitation issues that are expected to incur with respect to the ultimate power scaling of fiber lasers, such as efficiency degradation, thermal hazard, and system-instability growth in fiber lasers, and various relevant methods to alleviate the aforementioned issues. This discussion includes fiber nonlinear effects, fiber damage, and modal-instability issues, which become more significant as the power level is scaled up. In addition, we also review beam-combining techniques, which are currently receiving a lot of attention as an alternative solution to the power-scaling limitation of high-power fiber lasers. In particular, we focus more on the discussion of the schematics of a spectral beam-combining system and their individual requirements. Finally, we discuss prospects for the future development of fiber laser technologies, for them to leap forward from where they are now, and to continue to advance in terms of their power scalability.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.