<?xml version="1.0" encoding="UTF-8"?>
<feed xmlns="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <title>DSpace Collection:</title>
  <link rel="alternate" href="http://dspace.cityu.edu.hk:80/handle/2031/747" />
  <subtitle />
  <id>http://dspace.cityu.edu.hk:80/handle/2031/747</id>
  <updated>2013-05-31T18:28:09Z</updated>
  <dc:date>2013-05-31T18:28:09Z</dc:date>
  <entry>
    <title>Performance analysis of Multi-channel Aloha (MCA) and Multi-channel Slotted Aloha (MCSA) schemes for optical fiber network</title>
    <link rel="alternate" href="http://dspace.cityu.edu.hk:80/handle/2031/6629" />
    <author>
      <name>Yuen, Man Hon ( 袁文瀚)</name>
    </author>
    <id>http://dspace.cityu.edu.hk:80/handle/2031/6629</id>
    <updated>2012-08-07T07:47:52Z</updated>
    <published>2011-01-01T00:00:00Z</published>
    <summary type="text">Title: Performance analysis of Multi-channel Aloha (MCA) and Multi-channel Slotted Aloha (MCSA) schemes for optical fiber network
Authors: Yuen, Man Hon ( 袁文瀚)
Abstract: ﻿Advances in optical fiber technology have facilitated transmission of large volumes 
of data at different wavelengths in the same fiber network using wavelength division 
multiplexing (WDM). We are interested in analyzing the performance of a Multi-channel Multiple Access Scheme for WDM optical fiber networks. In this thesis, we propose to use the Aloha scheme as the multiaccess protocol because optical fiber 
provides huge volumes of bandwidth that are often far larger than user requirements. 
The advantage of low efficiency Aloha schemes is their low cost and simple structure. 
Performance of single-channel and multi-channel Aloha and Slotted Aloha schemes 
with variable packet length is analyzed using a simple closed-loop model. The analytical model assumes a perfect environment where transmission can fail only if 
packet collision occurs. Analysis parameters include collision probability, throughput, packet rate, and end-to-end delay. Analysis results are verified by some discrete 
C event simulation results supported at 95% confidence interval (C.I.). Most of the 
analysis results are very close to those produced by simulation programs employing 
binary exponential backoff (16 windows size) as the retransmission scheme. 
Analysis and simulation results have provided many interesting findings. The external packet arrival rate or packet rate is proportional to the number of available 
channels with the same collision probability. Collision probability decreases as the 
number of available channels increases, if external packet arrival rate remains constant. Several states are defined such as steep period, turning period and flat period 
to describe the shape of these curves. The middle point of the turning period is 
recommended as the most effective choice because the performance improvement is 
not great after this point and the required number of channels is not large. Although 
these values vary in different situations, most of them are about 0:15, a little bit 
smaller than the "threshold collision probability" of 0:2. 
Throughput increases along with increase in external packet arrival rate until it 
reaches the maximum point. After the maximum point, the trend changes from 
increasing to decreasing and finally it becomes zero. Throughput is proportional to the number of available channels while the utilization improves slightly or remains 
the same after the number of available channels is increased. 
End-to-end delay increases slowly until the external packet arrival rate reaches the 
corresponding arrival rate of "threshold collision probability" and then there is a 
steep upward slope after the arrival rate exceeds the threshold value. 
Binary exponential backoff schemes have been shown to have better performance 
than fixed retransmission probability approach and random selection of one interval for a fixed period approach. Either "BEB16" or "BEB32" are suggested to be 
used as retransmission schemes of multi-channel Aloha and Slotted Aloha schemes 
because their overall performance in simulation are the best and close to analysis 
results. In addition, Aloha scheme is proposed to be applied in short average packet 
length distribution while long average packet length distribution must choose Slotted Aloha scheme. For medium average packet length distribution, both schemes 
are applicable; the selection criterion depends on the cost or the performance.
Notes: CityU Call Number: TK5103.592.F52 Y83 2011; xx, 160 leaves : ill. (some col.)   30 cm.; Thesis (M.Phil.)--City University of Hong Kong, 2011.; Includes bibliographical references (leaves 116-118)</summary>
    <dc:date>2011-01-01T00:00:00Z</dc:date>
  </entry>
  <entry>
    <title>Optical properties and applications of surfactant modified deoxyribonucleic acid complex</title>
    <link rel="alternate" href="http://dspace.cityu.edu.hk:80/handle/2031/6628" />
    <author>
      <name>Yang, Xin ( 仰欣)</name>
    </author>
    <id>http://dspace.cityu.edu.hk:80/handle/2031/6628</id>
    <updated>2012-08-07T07:47:50Z</updated>
    <published>2010-01-01T00:00:00Z</published>
    <summary type="text">Title: Optical properties and applications of surfactant modified deoxyribonucleic acid complex
Authors: Yang, Xin ( 仰欣)
Abstract: ﻿Deoxyribonucleic acid (DNA) was modified with cationic surfactant hexadecyltrimethyl-ammonium chloride (CTMA), and the resultant DNA-CTMA complex was dissolved in an organic solvent. DNA-CTMA thin films, with thicknesses ranging from a few hundred nanometers to a few hundred microns, were obtained by both spin-coating and casting methods. DNA-CTMA channel waveguides were fabricated by two lithographic methods, namely, photolithography assisted reactive ion beam etching (RIE) and direct electron beam lithography (EBL). For the former method, the RIE conditions were optimized using a L16 (45) array based on Taguchi design, and the optimum parameters were obtained using analysis of variables (ANOVA). For the latter method, DNA-CTMA was found to behave as either a positive or a negative resist, depending on the developer. Dose tests were done to determine the optimum exposure and development conditions. EBL approach was also applied to fabricate DNA-CTMA microring resonators with different radii and gap widths. 
Optical properties of the thin films, channel waveguides and microring resonators were characterized. Refractive index was measured to be ~1.5, whilst material absorption losses were 0.28dB/cm, 0.05dB/cm, 0.43dB/cm and 0.56dB/cm at 0.63µm, 0.85µm, 1.31µm and 1.55µm. Propagation losses of DNA-CTMA channel waveguides were estimated by a cut back method, which were ~0.7dB/cm, ~0.6dB/cm and ~0.5dB/cm at 1.55µm, 1.31µm and 0.63µm. The waveguide surface profiles were examined by an atomic force microscope (AFM) and a scanning electron microscope (SEM), which showed a smooth surface with roughness of ~1nm, and a high anisotropy with undercut ~0.5µm. Transmission spectra of DNA-CTMA microring resonators were captured from an optical spectrum analyzer (OSA). The spectra were theoretically analyzed and compared, from which the major ring parameters like coupling efficiency and round trip loss were extracted. Finally, a temperature sensor was demonstrated based on the microring resonator, with a sensitivity of ~0.1nm/°C. 
This work explores three dimensional optical waveguides and microring resonators for the first time. The data presented in this thesis demonstrates that DNA-CTMA has characteristics superior to many traditional organic materials, indicating it is a promising material for future optical device applications.
Notes: CityU Call Number: TA1660 .Y36 2010; xvi, 142 leaves : ill.   30 cm.; Thesis (M.Phil.)--City University of Hong Kong, 2010.; Includes bibliographical references.</summary>
    <dc:date>2010-01-01T00:00:00Z</dc:date>
  </entry>
  <entry>
    <title>Self-organizing map : visualization and data handling</title>
    <link rel="alternate" href="http://dspace.cityu.edu.hk:80/handle/2031/6626" />
    <author>
      <name>Xu, Yang ( 徐楊)</name>
    </author>
    <id>http://dspace.cityu.edu.hk:80/handle/2031/6626</id>
    <updated>2012-08-07T07:47:46Z</updated>
    <published>2010-01-01T00:00:00Z</published>
    <summary type="text">Title: Self-organizing map : visualization and data handling
Authors: Xu, Yang ( 徐楊)
Abstract: ﻿Data visualization, which is the graphical presentation of data information, 
has been widely applied in industrial areas, e.g. signal compression, pattern 
recognition, image processing, etc. Self-Organizing Map (SOM), a widely used 
visualization method proposed by Kohonen, is an unsupervised learning network 
to visualize high-dimensional data in a low-dimensional map. SOM is able to 
present the data topology by assigning each datum to a neuron with the highest 
similarity, so that the data with similar features are mapped onto adjacent 
neurons. 
The traditional SOM uses the uniform grid map. Thus, some input data are 
projected onto the same neuron. Apparently, it is not an effective way for 
preserving the data relationship between clusters or within one cluster. And 
pre-defining the map size is another disadvantage of SOM. In this thesis, a new 
algorithm named Polar Self-Organizing Map (PolSOM) is proposed. PolSOM is 
constructed on a 2-D polar map with two variables, radius and angle, which 
represent data weight and feature respectively. Compared with the traditional 
algorithms which project data on a Cartesian map by using Euclidian distance as 
the only variable, PolSOM can manifest the precise data topology, and obtain the 
intra cluster density and inter cluster density by a new clustering criterion, 
synthetical cluster density (SCD). In PolSOM, not only similar data are grouped together, data characteristics are reflected by their positions on the map. 
PolSOM, in fact, adopts the conventional hard assignment for the training 
process. In this thesis, a new variant of PolSOM algorithm, named Probabilistic 
Polar Self-Organizing Map (PPoSOM), is proposed. Instead of using the hard 
assignment, PPoSOM employs the soft assignment that the assignment of a 
datum to a neuron is based on a probabilistic function. It is developed to enhance 
the visualization performance. It is worth noting that the obtained principled 
weight-updating rule of PPoSOM makes the visualization performance, 
measured by synthetical cluster density (SCD), greatly improve. 
SOM usually considers the whole data set in one go, whereas the 
representative data are not well utilized. The learning process is found to be rigid 
and time-consuming when one is dealing with large data sets. In this thesis, we 
propose to apply density based data reduction method as pre-processing, which 
means that the proposed method extracts representative data preliminarily for the 
SOM training. This method is found to be particularly useful in terms of reducing 
the overall computational time. 
Finally, an interesting application on World University Rankings is included 
in this thesis. A new perspective on studying the nature of different universities is 
introduced. SOM is employed to provide estimates for the missing scores of 
some universities. Principal component analysis (PCA) is used to analyze the 
underlying nature of each university. The results and analyses show that we can 
effectively determine the underlying characteristics of a university by studying how its ranking varies with the change of weights on different features. Our 
study shows that the proposed approach is more effective than simply relying on 
a linear weighted sum ranking system.
Notes: CityU Call Number: QA76.87 .X87 2010; xii, 105 leaves : ill.   30 cm.; Thesis (M.Phil.)--City University of Hong Kong, 2010.; Includes bibliographical references (leaves 99-105)</summary>
    <dc:date>2010-01-01T00:00:00Z</dc:date>
  </entry>
  <entry>
    <title>Enhanced stroke-based Chinese input methods for mobile devices</title>
    <link rel="alternate" href="http://dspace.cityu.edu.hk:80/handle/2031/6622" />
    <author>
      <name>Wong, Chi Kwan ( 王志君)</name>
    </author>
    <id>http://dspace.cityu.edu.hk:80/handle/2031/6622</id>
    <updated>2012-08-07T07:47:37Z</updated>
    <published>2011-01-01T00:00:00Z</published>
    <summary type="text">Title: Enhanced stroke-based Chinese input methods for mobile devices
Authors: Wong, Chi Kwan ( 王志君)
Abstract: ﻿During the last three and half decades, thousands of Chinese input methods for computing devices have been developed. Nevertheless, people are still looking for better Chinese input methods in terms of ease of use and memorization, high input speed and improved virtual keyboard implementation on mobile devices. The conventional well-known stroke-based Chinese input method using only five basic stroke types could achieve low learning curve and small keypad implementation, but its input speed is limited by the complexity of Chinese characters. To tackle this problem, simplified character and phrase encoding methods using (3+3) rules are proposed in this study. It is called G6 (diG-6). G6 only takes the first 3 strokes and the last 3 strokes of a character to enhance the performance of conventional stroke-based Chinese input methods. To further enhance the input speed, phrases are encoded with the first 3 strokes of the first character and the first 3 strokes of the last character of the phrases. Experimental results show that the G6 can improve ALIC (Average Length of Input Codes), HRFC (Hit Rate of First Character) and KSPC (Key Strokes Per Character) as compared with conventional T9 stroke-based Chinese input method. G6 has successfully implemented on MS-Windows, Apple OSX and Google Android platforms using QWERTY keyboard and virtual keypad. 
 
With the advancements of the electronic technology, inputting Chinese on mobile devices, especially touchscreen smartphones, become accustomed. Due to the popularity of touchscreen devices, the physical keyboards are being replaced by finger-operated virtual keyboards. With more devices using virtual keyboard, conventional Chinese input methods became adoptable for mobile devices. Most of the conventional Chinese input methods, however, are designed for QWERTY keyboard. The limited screen size of mobile devices incurs the input keys become tiny. It causes relatively high typo error rate that requires additional keystrokes to fix the errors. With the advancement of touchscreen technology, texting on touchscreen devices is no longer limited by finger tapping; it can also be performed by finger gestures. In this study, two novel dynamic keypad designs using finger gestures for character or radical selections are proposed to enhance the characters searching and input speed of the stroke-based Chinese input on touchscreen smartphones. The first approach is called DCK (Dynamic Candidate Keypad), in which eight frequently used Chinese characters are dynamically displayed on each of five stroke keys. These character candidates are changing with the input strokes and these characters can be directly selected by one of eight unidirectional finger gestures. Experimental results show that the proposed DCK can significantly enhance the top most 600 frequently used Chinese characters searching and input using stroke-based input method. The second approach, instead of candidate character selection, it applies dynamic keypad design for radical-stroke (RS) based Chinese character input using a dynamic radical keypad (DRK). In which commonly used Chinese radicals are dynamically displayed on the stroke keys of the DRK and similarly these radicals can be selected by unidirectional finger gestures. With the use of this new RS-DRK design, lot of radicals can be displayed and selected with a finger gesture motion and a few taps. It improves the input speed of complex Chinese characters that start or end with frequently used radicals. Experimental results demonstrated that RS-DRK could improve the KSPC, Total Error Rate and Character Per Minute for inexperienced and experienced users of RS-DRK as compared with conventional stroke-based Chinese input method.
Notes: CityU Call Number: PL1074.5 .W66 2011; xii, 105 leaves : ill. (some col.)   30 cm.; Thesis (M.Phil.)--City University of Hong Kong, 2011.; Includes bibliographical references (leaves 101-104)</summary>
    <dc:date>2011-01-01T00:00:00Z</dc:date>
  </entry>
</feed>

