Search

Existing video streaming protocols get poor QoE (Quality of Experience) especially when network bandwidth is drastically fluctuated due to handoff events over heterogeneous networks where they provide different bandwidths and network conditions. This paper proposes a novel QoE guaranteed mobile video streaming scheme over heterogeneous wireless networks: MOSQUITO. MOSQUITO can dynamically adjust streaming bit rate according to the network condition and also maintain fixed quantity of buffered AV (Audio-Video) data for the stable QoE level. MOSQUITO is implemented over the up-to-date Google Android mobile device. Simulation results show that that QoE in mobile video streaming service over heterogeneous wireless networks can be guaranteed by carefully predicting the network dynamics. MOSQUITO is now commercially deployed for over 5 million LTE subscribers of a south Korea telco, LG U+.

In this In this study, in order to develop low-temperature sintering ceramics for haptic piezoelectric actuator, Pb(Mg, W)O 3 -Pb(Ni, Nb)O 3 -Pb(Zr, Ti)O 3 (abbreviated as PMW-PNN-PZT) ceramics were fabricated using Li 2 CO 3 and CaCO 3 as sintering aids and their dielectric and piezoelectric properties were investigated with the amount of BiFeO 3 (abbreviated as BF) substitution. PMW-PNN-PZT composition ceramics could be sintered up to 920□ by adding sintering aids. As the amount of BF substitution was increased, the density, electromechanical coupling factor(kp), dielectric constant(er) and piezoelectric constant(d 33 ) showed the maximum value at each of sintering temperature. At the sintering temperature of 920□, the electromechanical coupling factor(kp), dielectric constant(sr) and piezoelectric constant(d 33 ) of 0.75 mol% substitution BF specimen showed the optimum values of 0.661, 2401 and 597 pC/N, respectively for piezoelectric haptic actuator application.

Lin and Chan proposed reversible secret sharing scheme in 2010. The advantages of this scheme are as follows: a low distortion, high capacity of shadow images and usage of the reversible method. However, this scheme has some problems. First, the number of participants is limited because of modulus prime number m. Second, the overflow can occur by additional operations in the embedding procedure. Finally, the coefficient of (t-1)-th degree polynomial can be zero due to the property of the cover image pixel. The proposed scheme solves these problems of the Lin and Chan's scheme. In order to solve these problems, a new polynomial technique are utilized in the proposed scheme. In the experimental results, PSNR of their scheme is decreased with the increase of embedding capacity. However, in the proposed scheme, even if the embedding capacity increase, PSNR value of about 45dB or more is maintained uniformly.

In this paper, we propose a texture feature-based language identification using Gabor and MDLC (multi-lag directional local correlation) features. In the proposed method, for a test image, Gabor magnitude images are first obtained by Gabor transform and magnitude operator and MDLC images by MDLC operator. Moments for the Gabor magnitude and MDLC images are then computed and fused into a feature vector. WPCA (whitened principal component analysis) finally searches one of training feature vectors most similar to the test feature vector. Experimental results show that the proposed method yields excellent language identification even with low feature dimension due to a well-matched fusion of Gabor and MDLC features.

We report a novel method of growing crystalline silicon layers by plasma enhanced chemical vapor deposition (PECVD) at 340° for solar cell application. Flash lamp annealing method was used for the seed formation on glass [1]. We optimized the precursor a-Si thickness to achieve 100% crystalline volume fraction with a large grain > 10 um. The proper surface treatment was found to be essential for the growth of an epitaxial layer on a poly-Si seed layer.

A microwave photonic (MWP) filter based on fiberoptic delay line has been proposed and experimentally demonstrated. By changing time-delay difference and coefficient of each wavelength signal by using fiber-optic delay lines and an electro-optic modulator, respectively, a bandpass filter or a notch filter can be implemented.

This paper proposes an adaptive edge-preserving smoothing and detail enhancement for video preprocessing of H.263. The proposed method consists of two steps: edge preserving smoothing using a motion-adaptive bilateral filter and detail enhancement using an adaptive high-boost filter. In the first step, the motion-adaptive bilateral filter preserves edges in region of interest (ROI) and smoothes noise in non-ROI, so that it can make video coder be more efficient under the restricted bitrate. Classification of ROI/non-ROI is done by simple thresholding of the absolute difference image between two adjacent frames. In the second step, an adaptive high-boost filter is used to enhance spatial edges of the decoded image. The bilateral filter and high-boost filter require user parameters according to the amount of motion and edgeness, respectively, and perform proper filtering with adaptive parameters. Experimental results show the effectiveness of the proposed method in terms of the visual quality and the peak signal-to-noise ratio.

Previous games including a scrolling shooter game lacked an emotional mechanism according to a personality of character. So characters could only express a simple and short emotion. If characters have emotions that is specialized their personality, they will show emotional response according to their personality. This effect will give emotional stimulus to players, so players' satisfaction and interactivity between characters and players will be improved. Therefore this paper purposes researching and developing an artificial emotion which can express emotions according to situations and character's personality in a scrolling shooter game. We will suggest three methods for our artificial emotions. First one is an emotion analyzing method which receives emotional stimulus and analyzes what kind of emotions have induced. Second one is an emotion controlling method which controls various emotions organically according to time and characters' personality. Last one is an emotion deciding method which decides emotional responses and actions according to emotions of present We made a scrolling shooter game based on an existing scrolling shooter game, and made another version which is applied our artificial emotion for estimating player's satisfaction And we compared and estimated two versions through a questionnaire.

This paper presents a novel QoS negotiable service framework (QNSF) to provide QoS based multi-room application services. The QNSF provides service bandwidth guaranteed services as well as priority based QoS services on home server based Ethernet networks. The QNSF works on Ethernet networks and legacy network devices. Although the QNSF can support priority based QoS services, in order to provide high quality multimedia services from subscriber networks, our scheme supports the service bandwidth guaranteed services because different home applications from multi-rooms compete for limited bandwidth resources. When application session setup begins, our QNSF negotiates a QoS service level of applications through QoS client APIs. Application services can perform any number of dynamic QoS negotiations before the service starts. Thus, the QNSF can provide well adapted QoS providing service to home multimedia applications. The QNSF consists of some functional blocks. A QoS client module defines QoS negotiation APIs for home service applications that require QoS negotiations. QoS proxy, QoS manager, and traffic monitor modules support service based QoS provision functions like dynamic bandwidth negotiation, QoS admission control, service profile based rate control, and adaptive parameter mapping. Packet classifier, class marker, and best effort traffic controller modules enable dynamic input traffic classification and support policy based rate controls. Using dynamic resource reservation and QoS negotiations, our scheme provides efficient QoS guaranteed services for the multimedia traffics from subscribe networks, such as high quality multimedia service. Our prototype implementation shows the feasibility of the QNSF for the provisioning QoS based application services in multi-room home networks

The home network has been revitalized due to the emergence of the broadband convergence network and the inter-networked home devices. The home gateway, which can be identified as a major component of the home network, enables to interconnect various home devices to one another as well as the Internet. Also, as the number of multimedia services, which requires high bandwidth and real-time support, increase users demand on the guaranteed quality for the services in the home network. In order to satisfy the user requirements, the Internet Service Providers (ISPs) have been providing a QoS mechanism for the home network services. However, the ISPs can not guarantee the end-to-end QoS from source to destination due to the misleading fact that the home gateway is recognized as an end terminal in terms of the ISP's viewpoint. Thus the QoS guaranteed mechanism for the remaining path, from the home gateway to the home devices, should be considered within the home network domain. In this paper, we propose a novel mechanism to guarantee the home network QoS for the services based on multicast traffics such as the IPTV or multilateral video conferencing. In order to meet the user requirements on the home network QoS, the proposed mechanism manages resources, classifies the traffic and also executes call admission controls for the service requested in the home network by utilizing the IGMP snooping scheme. Through the experiments of implementing the multicast based IPTV services, we will verify the proposed mechanism.

In this paper, we discuss about the experimental results of camera calibration and pose estimation for the application to vehicle's wheel alignment system using single camera. At first, we define a camera model for the camera system considered. For this model which consists of a linear part (ideal pinhole camera model) and a nonlinear part (lens distortion model), we drive a systematic and iterative calibration algorithm and pose estimation algorithm. In addition, we develop a coordinate matching algorithm which calculates matched sequence of order in both extracted image coordinate and object coordinate for non-interactive calibration and pose estimation. Based on the results of calibration, we also develop a camera simulator to confirm the results of calibration and compare the results of simulations with those of experiments in pose estimation

In this paper, we will discuss the implementation and performance issue of the IPSec VPN for the home gateway. We have developed the hardware accelerated IPSec protocol over the home gateway which is providing IPv4/v6 dual stack, IEEE1394 interconnection, IP multicast based HDTV (high definition TV), VoIP, VDSL and real time visitor confirming capability

This paper proposes a novel routing protocol to reduce routing update overhead in personal area network mobility (PANEMO). In order to reduce routing update overhead, this paper introduces the notion of multi-level region. Mobile routers (MRs) exchange link state information with their neighbor MRs with a frequency depending upon hop count to the destination. From updated link state information, MRs construct the topology map of the entire network and compute optimal routes.

This paper presents unsupervised moving object segmentation and recognition method for intelligent transportation systems (ITS). The presented method consists of three procedures: First, the object detection procedure, in which the rough positions of moving objects in an image sequence are determined using an adaptive thresholding method; Second, the object segmentation procedure, in which pixels that have similar intensity and motion information segments are grouped using a weighted k-means clustering algorithm to the binary motion mask obtained in the object detection. Finally, the object recognition procedure, in which a neural network is used to recognize whether the segmented objects are vehicles, humans, or other objects. The experimental results demonstrate robustness not only in variations of luminance conditions, but also for occlusions among multiple moving objects.

In a flow control system with a fan, the variable speed drive is becoming popular because of its cost-effectiveness compared with conventional vane or damper flow control systems. But, for the large system which requires the parallel operation of several fans, it cannot always be said that driving all the fans with variable speed is the best or most cost-effective choice. In this regard, a variable speed drive fan and a constant speed drive fan with vane control under conditions of parallel operation are dealt with. The total energy consumed in the vane-speed parallel flow control system depends on the operating points of the fan speed and the vane opening, and three will exist an optimal combination of the speed of one and the vane opening of the other that minimizes the total power consumption. This optimal combination is derived from a mathematical model and tested on actual fans for a generating boiler system. The results confirm the validity of the theoretical analysis. >

Vector controlled drives of a PMSM (permanent magnet synchronous motor) are proposed without position and speed sensor. The conventional pole position and speed sensors are eliminated by means of estimation using adaptive identification, for which stator voltages and currents are measured. To realize the proposed algorithm, a d-q current detecting method, which is independent of the rotor speed and position, is proposed. Simulation and experiment show that the proposed sensorless algorithm has good adaptation with load condition variation. Simulation results show that the position error at the steady state is within one electrical degree. In the experiment, however, the error becomes larger than that value, because of the sensing error and the long sampling time. >

A current regulator for the resonant inverter, which chooses the nearest available voltage vector to the predicted desired voltage vector is proposed. It is proved that the proposed regulator minimizes RMS current error. The performance of the regulator is tested by computer simulations and compared with that of a conventional pulse density modulation (PDM) regulator. The effects of the parameter and estimation error in the back EMF are considered in the computer simulation. The implementation of the regulator using a digital signal processor and its experimental results are described. The computer simulation and experimental results reveal improved responses compared with a conventional PDM regulator. >

A control technique for active power filters is proposed which not only eliminates the harmonic current but also controls the reactive power at the AC side of a PWM (pulsewidth-modulated) inverter-induction drive system. Injecting the proposed PWM current enables the harmonic components of orders not greater than the pulse number per half-cycle to be removed completely. It also enables the input fundamental power factor to become unity. Theoretical investigations are performed to evaluate the performance of the proposed control technique. >

A load commutated current source inverter (LCCSI) induction motor drive system employing a novel DC-side forced commutation circuit for machine startup is proposed. To avoid the adverse effects of the induction motor, a solution is proposed, i.e., using a special kind of one-notch current pulse width modulation (PWM) fully utilizing the proposed commutation circuit. Along with compensating the capacitive current, the direct vector control referred to the rotor flux fixed reference frame without speed sensor is used to decouple torque and flux generation currents and to overcome the problem of instability caused by the output capacitor. The availability of the proposed circuit for a high-power drive system is verified by computer simulation for a 600 hp induction motor. >

In this paper, observer-based state feedback speed controller for torsional vibration suppression of a 2-mass system is proposed. Because of the difficulty of measuring the load speed and the shaft torque in real plant, the reduced observer is used to estimate not only the unknown states but also the load torque disturbance. Simulations and experiments carried out to verify this algorithm in comparison with conventional PI controller. This controller shows better command following characteristic without torsional vibration. The speed response to the rapid load torque change is also examined to show the disturbance rejection characteristic which is specially important in industrial applications such as rolling mill drives. At the moment of impact load, the speed drop is reduced and the recovery time is shortened remarkably. >

A new generator excitation system using a boost-buck chopper as a kind of static exciter is proposed to overcome the lack of field forcing capability of the bus fed exciter under the input line fault condition. It increases or maintains the generator field current by boosting the field voltage in the case of the input AC line voltage drop during and immediately after a fault. The validity of the proposed excitation system is verified with the computer simulation. The generator stability according to different exciters is tested using a commercial software package-CYME. The simulation results of the stability analysis on the generator with the proposed exciter is better than that of the bus fed exciter. This boost-buck chopper exciter can be simply implemented and controlled by the modern power electronics technology.

In the power range of above 1 MW, industrial AC drive load-commutated current source inverters (LCCSI) are controlled by simple methods like the combination of V/f ratio constant control at low frequency and V/f ratio constant-angle control at high frequency. This gives a good steady-state but poor dynamic performance. This paper describes a drive control of LCCSI-fed induction motors which is based on the principle of field oriented control. Along with taking into account the output capacitor's effect, the direct field oriented control referred to rotor flux fixed reference frame without a speed sensor is applied to decouple torque and flux generating currents and to overcome the problem of instability caused by the output capacitor. This control scheme is implemented by a 32 bit floating point Digital Signal Processor (DSP) chip-TMS320C30. The availability of this control scheme is verified by computer simulation and experimental results for a laboratory LCCSI-induction motor (10 HP) system. >

A comparison of the different types of current regulator for an induction motor drive system fed by IGBT inverter is presented. The two parts of current regulation are considered: current error compensation and voltage modulation. The RMS value of current error is chosen as a performance index. The characteristics of hysteresis, synchronous PI, predictive, deadbeat and state feedback controllers are analyzed and the steady-state performances and transient responses of current regulation are compared through simulation and experimental results. >

This paper describes the implementation of self-tuning integral-proportional (IP) speed controller and predictive current controller for a rolling mill DC motor drive system, which uses the Texas Instrument TMS 320C30 digital signal processor (DSP) chip. With the predictive current control algorithm, the response of the system is improved significantly and the excellent dynamic response is also guaranteed over the whole operating conditions including discontinuous current and current reversal. Digital simulation and experimental results show that proposed controllers possess excellent adaptation capability as well as transient recovery under load changes. >

In this paper, a novel strategy of current regulation of a field-orientation induction motor drive using the GTO inverter is proposed, which is based on a multivariable state feedback control with an integrator. The controller is designed by pole placement technique of multivariable system regulation theory. In addition, feedforward control is included in the control laws to improve transient responses. For the linear state feedback control, the rotor flux is separated as a disturbance and is estimated with a reduced-order state observer. A good steady-state performance is obtained by means of an integral compensation and symmetrical space voltage vector PWM, and fast transient response is also feasible since the required voltage is directly calculated from the feedforward control. Also, it is demonstrated that an instantaneous current control without the calculation time delay can be implemented. The excellence of the proposed current regulation scheme is verified by experiments for the GTO inverter-fed induction motor drive system controlled by a DSP-TMS320C30. >