Browsing by Author "Chang, WH"

Now showing 1 - 9 of 9

A high throughput context-based adaptive arithmetic codec for JPEG2000
(2002)
Ong, KK
;
Chang, WH
;
Tseng, YC
;
Lee, YS
;
Lee, CY
In JPEG2000 image compression standard, context-based arithmetic coding is adopted for improving compression rate. Efficient and high throughput codec: design is strongly required for handling high-resolution images. We propose an efficient architecture for context-based adaptive arithmetic coding, which can achieve high throughput performance. Based on parallel leading zeros detection and bit-stuffing handling, symbol can be encoded and decoded within one cycle. Therefore, throughput rate can be increased as high as codec operating frequency. The modified encoding and decoding flows are compatible with JPEG2000 standard. For 0.35u 1P4M CMOS technology, encoding and decoding rate can run up to 200Msymbol/sec and 175Msymbol/sec respectively. They are only cost 6.8K and 9.8K gate count. These high throughput performances can meet high-resolution real time application requirement.
A high throughput low cost context-based adaptive arithmetic codec for multiple standards
(2002)
Ong, KK
;
Chang, WH
;
Tseng, YC
;
Lee, YS
;
Lee, CY
For next generation image compression standard, context-based arithmetic coding is adopted for improving compression rate. Efficient and high throughput codec design is strongly required for handling high-resolution images. We propose an efficient codec architecture for context-based adaptive arithmetic coding, which exhibits low cost, low latency, and high throughput rate. In addition, it can be programmed for supporting multiple standards such as JPEG, JPEG2000, JBIG, and JBIG2 standards. It is exploiting three-pipeline stages architecture. Based on parallel leading zeros detection and bit-stuffing handling, symbol can be encoded and decoded within one cycle. Therefore, throughput rate can be increased as high as codec operating clock rate. For 0.35u 1P4M CMOS technology, both encoding and decoding rate can run up to 185Msymbol/sec. The AC codec only costs 12K gate count and 860um x 860um layout area. These performances can meet high-resolution real time application requirement.
A multiobjective hybrid genetic algorithm for the capacitated multipoint network design problem
(2000-06-01)
Lo, CC
;
Chang, WH
;
資訊管理與財務金融系註：原資管所+財金所
;
Department of Information Management and Finance
The capacitated multipoint network design problem (CMNDP) is NP-complete, In this paper, a hybrid genetic algorithm for CMNDP is proposed. The multiobjective hybrid genetic algorithm (MOHGA) differs from other genetic algorithms (GA's) mainly in its selection procedure. The concept of subpopulation is used in MOHGA, Four subpopulations are generated according to the elitism reservation strategy, the shifting Prufer vector, the stochastic universal sampling, and the complete random method, respectively, Mixing these four subpopulations produces the next generation population. The MOHGA can effectively search the feasible solution space due to population diversity, The MOHGA has been applied to CMNDP. By examining computational and analytical results, we notice that the MOHGA can find most nondominated solutions and is much more effective and efficient than other multiobjective GA's.
A new pre-etching pattern to determine < 110 > crystallographic orientation on both (100) and (110) silicon wafers
(2005-02-01)
Chang, WH
;
Huang, YC
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
Since the etching rate of anisotropic wet etching of silicon is highly orientation-dependent, it is important to determinate the crystallographic orientation before etching. In this paper, a new mask pattern based on the pre-etching process is designed to find out the <110> crystal orientation on both (100) and (110) silicon wafers observably and to align the subsequent masks more easily and conveniently. Because of a compensation of the quantization error due to the limitated scanning resolution of a mask generator and an optimum design, a pre-etching mask pattern with a resolution of 0.00625degrees and a span angle of +/-2degrees is achieved with a consumption area of 560 x 14,472 mum(2). To evaluate the precision of the misorientation angle determined by the pre-etching process based on the proposed pattern, both (100) and (110) wafers with 60 pre-etching patterns on each of them were etched. After statistic analysis, the precision of +/-0.0467degrees for (100) wafers and +/-0.025degrees for (110) wafers are achieved in our experiments which suggest the precision of determination of the crystallographic misorientation by the preetching process may be limited even though the resolution of the pre-etching mask pattern is high.
Electrochemical etching of n-type 6H-SiC without UV illumination
(2006-06-01)
Chang, WH
;
Schellin, B
;
Obermeier, E
;
Huang, YC
;
交大名義發表
Deep etching of n-type 6H-SiC using a two-step etching process has been studied. First, anodization of 6H-SiC in an HF electrolyte (2 wt.%) without ultraviolet light is applied to form a deep porous layer with the desired dimensions. Then, a thermal oxidation process is used to oxidize this porous layer. The oxidized layer is then removed in a concentrated HF solution. In the experiments, the etching parameters electrolyte concentration and current density are optimized in order to obtain a uniform pore size and hence, a smooth etched surface. After adjusting these parameters, the porous layer formation experiments are carried out at 20 degrees C in a 2 wt.% HF electrolyte using a current density of 50 mA/cm(2). The corresponding porous layer formation rate is about 1.1 mu m/min. To demonstrate the capabilities of this SiC bulk micromachining process, deep circular cavities are fabricated in n-type 6H-SiC substrates.
Micromachining of p-type 6H-SiC by electrochemical etching
(2004-04-15)
Chang, WH
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
In order to structure p-type single-crystal 6H-SiC wafers for micro-electro-mechanical systems (MEMS) applications, electrochemical etching at a constant current density is used. In this paper, first, a suitable mask with a Poly/SiO2/SiC scheme is used for a long-time electrochemical etching of p-type 6H-SiC in a dilute HF solution. Etching rates and etched profiles with different concentrations (1, 2, 4, 8%) of the HF solution and different current densities (50, 100, 150 mA/cm(2)) are investigated in order to optimise the etching process. An HF concentration between 2 and 4% and a current density between 50 and 100 mA/cm(2) are suggested to obtain a smooth etching surface and a uniform etching profile. With these process parameters, the etching rate is found to be between 0.40 and 0.95 mum/min. A circular structure with a depth of 200 mum is demonstrated for further MEMS applications. (C) 2003 Elsevier B.V. All rights reserved.
Novel cleaning solutions for polysilicon film post chemical mechanical polishing
(2000-07-01)
Pan, TM
;
Lei, TF
;
Chen, CC
;
Chao, TS
;
Liaw, MC
Novel cleaning solutions were developed for post-CMP process, surfactant tetra methyl ammonium hydroxide (TMAH) and/or chelating agent ethylene diamine tetra acetic acid (EDTA) were added into the diluted ammonium hydroxide (NH4OH+H2O) alkaline aqueous solution to enhance removal of metallic and organic contamination, From the experimental result, it is found that the particle and metal removal efficiency and the electrical characteristics are significantly improved for post-CMP cleaning.
Photoluminescence of colloidal CdSe/ZnS quantum dots under oxygen atmosphere
(2005-09-01)
Shu, GW
;
Lee, WZ
;
Shu, IJ
;
Shen, JL
;
Lin, JCA
The effects of oxygen versus vacuum ambients on colloidal CdSe/ZnS quantum dots (QDs) were studied using both continuous and time-resolved photoluminescence (PL) measurements. The PL intensities were found to be an order of magnitude higher in an oxygen atmosphere, which is explained by the passivation of surface defects by oxygen absorption. The decay of PL intensities can be best fitted by a biexponential function with lifetimes of approximately 1 ns; for the fast decay and approximately 10 ns; for the slow decay. Based on the emission-energy dependence of carrier lifetimes and of the amplitude ratio of the fast-decay component to the slow-decay component, we suggest that the fast and slow PL decay of colloidal CdSe/ZnS QDs is caused by the recombination of delocalized carriers in the internal core states and the localized carriers in the surface states, respectively.
Recombination dynamics of luminescence in colloidal CdSe/ZrS quantum dots
(2005-09-01)
Lee, WZ
;
Shu, GW
;
Wang, JS
;
Shen, JL
;
Lin, CA
Recombination dynamics of photoluminescence (PL) in colloidal CdSe/ZnS quantum dots (QDs) were studied using time-resolved PL measurements. The PL intensity shows a biexponential decay at 9 K, consisting of a fast component (similar to 1 ns) and a slow component (similar to 6.3 ns). Based on the emission-energy and temperature dependence of carrier lifetimes, we suggest that the fast and slow PL decay of colloidal CdSe/ZnS QDs originates from recombination of the delocalized carriers in the internal core states and the localized carriers at the interface, respectively.