Browsing by Author "Chen, JR"

Now showing 1 - 19 of 19

A distributed end-to-end rate control scheme for ABR service
(1998)
Chen, JR
;
Chen, YC
;
Chan, CT
;
資訊工程學系
;
Department of Computer Science
In ATM networks, the ABR service allows both the source and the destination to function simply, this means that we may need a certain complexity in the switch. To provide robust ABR service, policing and discarding functions are usually required. ABR rate control schemes such as UT [4] and OSU [5][7] can achieve high utilization, but they suffer the oscillation of transmission rate. We propose an End-to-End Rate Control Algorithm (EERCA), which performs the monitoring function in end systems, and achieves max-min fairness naturally. Our scheme relies on the round-robin discipline in combination with per-VC queuing for cell scheduling, it also covers the policing function. EERCA resolves the oscillation problem that is quite common in existing explicit rate schemes, as well as features very low cell loss ratio with much smaller buffer size.
A scalable architecture for differentiated services
(2002)
Lee, CL
;
Chen, JR
;
Chen, YC
;
交大名義發表
;
National Chiao Tung University
Differentiated services (DiffServ) model is a potential solution for providing quality of services (QoS) on the Internet. In this paper, we propose a scalable architecture that fulfills the design philosophy of the DiffServ model. Different from other existing architectures, our proposed architecture puts complexity on edge routers so that core routers do not need to support any specific function for service differentiation. Further, the proposed architecture only needs few additional functions in edge routers. Thus its deployment is simpler than most existing architectures. We also address how to achieve weighted proportional fairness solely by edge routers.
A simplified approach based on source control for ATM ABR service
(2001-08-15)
Lee, CL
;
Chen, YC
;
Chen, JR
;
資訊工程學系
;
Department of Computer Science
The available bit rate (ABR) service is designed to provide efficient support of data traffic in ATM networks. A variety of rate-based flow control approaches have been proposed to support ABR service. Most of these approaches improve both the fairness among active connections and the link utilization by putting more and more complexity into a switch. In this article, we propose a flow control approach in which part of rate-calculation work is moved from switches to end-systems. As a result, the proposed approach reduces the rate-calculation effort in the switch. This may reduce the switch complexity and thus the implementation cost. Furthermore, it mitigates the difficulty for setting the measurement interval in switches. as well as features lower and more stable queue occupancy in the network. Simulation results are presented to demonstrate that the proposed approach achieves better performance than ERICA +, a well-known switch algorithm. (C) 2001 Elsevier Science B.V. All rights reserved.
An A-trimmed skeleton for modeling the global shape of polygons
(2006-02-01)
Wu, MC
;
Chen, JR
;
Chuang, JH
;
資訊工程學系
;
工業工程與管理學系
A skeleton is a global shape descriptor that aims to characterize the general appearance of an object to facilitate object classification. Various ways for defining the skeleton of a polygon have been proposed in the literature. Yet, in some cases, these previously proposed skeletons may not intuitively resemble the original polygons and cannot be a good shape descriptor. This research proposes a new definition of skeletons for modeling polygons. The proposed skeleton is termed the area-trimmed (A-trimmed) skeleton. In modeling the global shape of polygons, the A-trimmed skeleton performs much better than other skeletons proposed in the literature. The A-trimmed skeleton can be applied in group technology and facilitate tasks such as punched-part design and fixture design.
An end-to-end flow control approach based on round trip time
(2000-09-30)
Chen, JR
;
Chen, YC
;
Lee, CL
;
資訊工程學系
;
Department of Computer Science
Traditional TCP Row control schemes, such as Tahoe and Reno, continuously increase the window size until possible packet loss is detected. These schemes generate congestion unnecessarily by themselves and utilize the buffer space inefficiently. Therefore, Vegas was proposed to accommodate the drawbacks. It uses the measured round-trip time (RTT) to obtain the information required for the flow control. in this work, we propose a Row control scheme, RTT-based TCP, which uses the RTT directly to control the window size adjustment. RTT-based TCP comprises four mechanisms: the modified exponential window increment, the RTT measurement, the queue occupancy estimation, and the quick queue-occupancy reduction. Simulation results show that RTT-based TCP outperforms Vegas in the aspects of throughput, fairness and efficiency of buffer usage. (C) 2000 Elsevier Science B.V. All rights reserved.
Automated information. mining on multimedia TV news archives
(2005)
Lai, PS
;
Cheng, SS
;
Sun, SY
;
Huang, TY
;
Su, JM
This paper addresses an integrated information mining techniques for multimedia TV-news archive. The utilizes techniques from the fields of acoustic, image, and video analysis, for information retrieval on news story title, newsman and scene identification. The goal is to construct a compact yet meaningful abstraction of broadcast news video, allowing users to browse through large amounts of data in a non-linear fashion with flexibility and efficiency. By using acoustic analysis, the system can classify video into news versus commercials, with 90% accuracy on a data set of 400 hours TV-news recorded off the air from July 2003 to August of 2004. By applying speaker identification and/or image detection techniques, each news stories can be segmented with an accuracy of 96%. On screen captions or subtitles are recognized by OCR techniques to produce the text title of each news stories. The extracted title words can be used to link or to navigate more related News contents on the WWW. In cooperation with facial and scene analysis and recognition techniques, OCR results can provide users with multimodality query for specific news stories. Some experimental results are presented and discussed for the system reliability and performance evaluation and comparison.
Design and analysis of queuing delay control for end-to-end rate control algorithm
(1999-10-01)
Chen, JR
;
Chen, YC
;
資訊工程學系
;
Department of Computer Science
distributed congestion avoidance scheme, the End-to-End Rate Control Algorithm (EERCA), has been proposed [5] for Available Bit Rate (ABR) service. In this work, we enhance EERCA by replacing the Queue Occupancy Reduction with the queuing delay control (QDC), which manipulates the queue occupancy more efficiently with less functional complexity. Furthermore, it alleviates per-VC accounting requirement and thus reduces the complexity of a switch. From both the analysis and the simulation, it shows that the proposed virtual queue occupancy precisely reflects the variation of the actual queue occupancy. This enhanced EERCA features higher efficiency and more stable queue occupancy than the original algorithm.
Effects of isolation materials on facet formation for silicon selective epitaxial growth
(1997-10-20)
Tseng, HC
;
Chang, CY
;
Pan, FM
;
Chen, JR
;
Chen, LJ
Effects of isolation materials on facet formation for low temperature Si selective epitaxial growth (SEG) using Si2H6 have been demonstrated by ultrahigh vacuum-chemical molecular epitaxy (UHV-CME) system. The activation energy of the Si epitaxial growth is 47.35 Kcal/mol. The isolation material includes silicon nitride, wet oxide, and tetraethylorthosilicate (TEOS) oxide. Different isolation materials have different faceting behaviors for the Si epilayer. A silicon nitride mask and a wet oxide mask result in a preferred faceting plane of (111) and (311), respectively. Using a TEOS oxide mask, we obtain a low temperature Si SEG layer without facet and twin formation, The preferred faceting plane orientation for different isolation materials is ascribed to the difference in the interface free energy between the Si layer and the isolation material. (C) 1997 American Institute of Physics.
High-kappa Ir/TiTaO/TaN capacitors suitable for analog IC applications
(2005-07-01)
Chiang, KC
;
Huang, CC
;
Chin, A
;
Chen, WJ
;
McAlister, SP
We have developed novel high-k Ir/TiTaO/TaN capacitors which have high-capacitance density (10.3 fF/mu m(2)), small leakage current at 2 V (1.2 x 10(-8) A/cm(2)), and low voltage linearity of the capacitance (89 ppm/V-2). These excellent results meet the TTRS roadmap requirements for precision -analog capacitors for the year 2018. The good performance is due to the very high K (45) achieved in the TiTaO dielectric and the high work-function (5.2 eV) provided by the Ir electrode.
Improving data transport performance through SSCOP with rate-based flow control
(1999-12-01)
Chen, JR
;
Chen, YC
;
資訊工程學系
;
Department of Computer Science
This work tries to improve the effective throughput as well as to achieve better fairness of the transport protocol with efficient selective retransmission scheme. The commonly used transport layer protocol, transmission control protocol (TCP), has been developed for more than two decades, but its retransmission scheme is hard to be satisfactory. However, service specific connection oriented protocol (SSCOP) provides an excellent retransmission mechanism with poor flow control function. When the lower layer protocol is Asynchronous Transfer Mode, using SSCOP instead of TCP/IP over AA5 to provide transport layer service can specifically reduce the protocol overhead. We propose a scheme for SSCOP to enhance its flow control performance. This scheme combines the rate-based flow control with the queuing delay control. The simulation result shows that our enhanced SSCOP achieves both higher throughput and better fairness. It demonstrates that a proper flow control scheme with efficient selective retransmission is crucial for optimizing the performance of a transport protocol. (C)1999 Elsevier Science B.V. All rights reserved.
Numerical analysis on hot spot in reactive chemical storage
(1998)
Sheu, LJ
;
Lin, JD
;
Chen, JR
;
機械工程學系
;
Department of Mechanical Engineering
In an exothermic reactive chemical storage, a thermal explosion can occur when the heat generation within the storage is greater than heat removal from the storage. In a solid system, the existence of an inflection point in the temperature versus time curve is often used as a criterion for thermal explosion. In this study the applicability of this criterion to a fluid storage is examined by investigating the hot spot in the system. Transient natural convection of an exothermically reactive fluid in the vertical cylindrical storage tank with isothermal walls is investigated numerically. The axisymmetric 2-D Navier-Stokes equations governing the now fields are reduced by introducing the stream function-vorticity formulation and solved by the alternating-direction-implicit (ADI) technique. The reactive heat sources are represented by a zeroth order rate expression with a Arrhenius type rate constant. It is found that the hot spot is no longer located in the center of the tank due to the buoyancy effect. In a particular range of Rayleigh number and Frank-Kamenetskii number, the hot spot is found to move periodically. A stable steady regime, stable oscillatory regime and thermal explosion regime can be specified based on the dependence of the critical Frank-Kamenetskii number on the Rayleigh number. In the stable oscillatory regime multiple inflections in temperature versus time are found. Therefore, the existing criterion which defines thermal explosion by the appearance of an inflection in the temperature versus time curve is no longer applicable in reactive fluid systems. The results of nonconvective systems do, however, provide a preliminary and conservative estimate for thermal explosion of a reactive storage.
Performance evaluation of an end-to-end rate control approach for ABR services
(1998-12-01)
Chen, JR
;
Chen, YC
;
Chan, CT
;
資訊工程學系
;
Department of Computer Science
In this work, we propose an End-to-End Rate Control Approach (EERCA) for congestion avoidance in Available Bit Rate (ABR) service on Asynchronous Transmit Mode (ATM) networks. In our approach: the network estimates the number of cells stored in the switch for each VC. The source generates a specific traffic pattern, then a proper explicit rate can be derived based on the received traffic pattern at the destination. This approach is designed to reduce the rate calculation effort in the switch as well as to avoid the complexity in setting the monitoring-interval. EERCA features higher efficiency, higher utilization, more stable queue occupancy shorter transient response time, and better fairness compared with existed schemes.
Performance evaluation of an end-to-end rate control approach for ABR services
(1998-12-01)
Chen, JR
;
Chen, YC
;
Chan, CT
;
資訊工程學系
;
Department of Computer Science
In this work, we propose an End-to-End Rate Control Approach (EERCA) for congestion avoidance in Available Bit Rate (ABR) service on Asynchronous Transmit Mode (ATM) networks. In our approach: the network estimates the number of cells stored in the switch for each VC. The source generates a specific traffic pattern, then a proper explicit rate can be derived based on the received traffic pattern at the destination. This approach is designed to reduce the rate calculation effort in the switch as well as to avoid the complexity in setting the monitoring-interval. EERCA features higher efficiency, higher utilization, more stable queue occupancy shorter transient response time, and better fairness compared with existed schemes.
Providing service differentiation in wireless networks
(2001-04-01)
Lee, CL
;
Chen, YC
;
Chen, JR
;
資訊工程學系
;
Department of Computer Science
Differentiated services: (Diffserv) model is one of the possible solutions for providing quality of service (QoS) on the Internet. Most existing approaches assume that the packet loss is an indication of network congestion and thus reduce the sending rates: of sources. For wireless networks, the assumption is not proper since packet losses may he caused by other reasons, such as fading and interference of the signal. Therefore, these approaches do not work well in wireless networks. In this paper, we propose an approach which is able to provide service differentiation in wireless environments. In our approach, the rate share of a connection is determined by the associated weight. By keeping a proper amount of extra data in the network, the proposed approach can achieve weighted proportional fairness, which can provide selective QoS without any particular support, from the network. We use the ns simulator to evaluate our approach. Simulation results show the validity of the proposed approach.
Pseudo-rate TCP: a congestion avoidance scheme with nearly optimized fairness and throughput
(1999-09-25)
Chen, JR
;
Chen, YC
;
資訊工程學系
;
Department of Computer Science
As the bandwidth requirement on the Internet grows, the efficiency of networking protocol becomes an important concern. This paper proposes a so-called pseudo-rate TCP, which features both better fairness and higher throughput than TCP Vegas by exponentially increasing the window size to a derived bound. The pseudo-rate TCP avoids the typical problem that sources with shorter round-trip times always get better chance in allocating the bandwidth. We discuss two key mechanisms designed for pseudo-rate TCP, and compare its performance with TCP Vegas through system simulation. (C) 1999 Elsevier Science B.V. All rights reserved.
Synthesis and physical properties of linear and branch phenoxies
(2001-12-01)
Chen, JR
;
Li, YS
;
Chang, FC
;
應用化學系
;
Department of Applied Chemistry
Linear and branch phenoxies have been synthesized by reacting difunctional epoxy and diphenol or triphenol monomer, using phosphonium salt as the catalyst. The reaction kinetics is analyzed by the ASTM E 698 method, which is suitable for a first order reaction system. The bulky structure of the bisphenol Z moiety retards the reaction; therefore, the reaction temperature required is higher- at 200 degreesC, compared to 190 degreesC for the bisphenol A based system. The introduction of a branch structure increases both molecular weight and the polydispersity of the phenoxy. A slightly branched structure tends to increase the distance between the polymer chains and results in a lower shear viscosity for easier processing. A higher branch results in light crosslinking and decreases the mobility of the polymer. The branch structure has a negligible effect on its glass transition temperature because the average distance between branches is substantially longer than the critical mainchain length for T-1's Brownian motion. The 100/1 (bisphenol/ trisphenol) system has the lowest viscosity, but its mechanical properties and the gas permeation properties are poorer than those of the others. The bisphenol Z based phenoxy shows superior mechanical properties, higher T-g, lower coefficient of thermal expansion, and lower gas permeation properties than the corresponding bisphenol A based phenoxy because of the steric hindrance structure of the bisphenol Z moiety.
Theoretical and experimental analysis on InAlGaAs/AlGaAs active region of 850-nm vertical-cavity surface-emitting lasers
(2006-01-01)
Chang, YA
;
Chen, JR
;
Kuo, HC
;
Kuo, YK
;
Wang, SC
In this study, the gain-carrier characteristics of In0.02Ga0.98As and InAlGaAs quantum wells (QWs) of variant In and Al compositions with an emission wavelength of 838 nm are theoretically investigated. More compressive strain, caused by higher In and Al compositions in InAlGaAs QW, is found to provide higher material gain, lower transparency carrier concentration, and transparency radiative current density over the temperature range of 25-95 degrees C. To improve the output characteristics and high-temperature performance of 850-nm verticalcavity surface-emitting laser (VCSEL), In(0.15)AI(0.08)Ga(0.77)As/ Al0.3Ga0.7As is utilized as the active region, and a high-bandgap 10-nm-thick Al-0.75 Ga0.25As electronic blocking layer is employed for the first time. The threshold current and slope efficiency of the VCSEL with A(10.7)SGa(0.25)As at 25 degrees C are 1.33 mA and 0.53 W/A, respectively. When this VCSEL is operated at an elevated temperature of 95 C, the increase in threshold current is less than 21 % and the decrease in slope efficiency is approximately 24.5%. A modulation bandwidth of 9.2 GHz biased at 4 mA is demonstrated.
Vegas plus: Improving the service fairness
(2000-05-01)
Chen, JR
;
Chen, YC
;
資訊工程學系
;
Department of Computer Science
TCP Vegas is a congestion avoidance scheme designed to prevent the periodic packet loss which occurs in traditional schemes. Since Vegas successfully avoids such packet loss, it achieves much higher throughput than TCP Reno. However, it does not concern the fairness among source-destination pairs with different round-trip times (RTT's). In this letter, we propose a different mechanism to adjust the window size, this allows TCP to provide much better fairness regardless the large variation of RTT's.
Very high-density (23 fF/mu m(2)) RF MIM capacitors using high-k TaTiO as the dielectric
(2005-10-01)
Chiang, KC
;
Lai, CH
;
Chin, A
;
Wang, TJ
;
Chiu, HF
A very high density of 23 fF/mu m(2) has been measured in RF metal-insulator-metal (MIM) capacitors which use high-kappa TaTiO as the dielectric. In addition, the devices show a small reduction of 1.8% in the capacitance, from 100 kHz to 10 GHz. Together with these characteristics the MIM capacitors show low leakage currents and a small voltage-dependence of capacitance at 1 GHz. These TaTiO MINI capacitors should be useful for precision RF circuits.