Volume 5, Issue 5, May 2014

A New Genetic Algorithm Applied to Inexact Graph Matching
Pages: 1-6 (6) | [Full Text] PDF (480K)
Le Dang Nguyen, Dac-Nhuong Le, Tran Thi Huong and Le Trong Vinh
Haiphong University, Haiphong, Vietnam
Hanoi University of Science, Vietnam National University, Hanoi, Vietnam

Abstract -
Graph matching is used for model-based pattern recognition of brain images, model design objects in a computer-aided design, machine learning, data mining, packet filtering, web phishing, etc. In this paper, we have proposed a new genetic algorithm for inextract graph matching with many types of graph such as undirected, directed, weighted, and labeled. The experimental results show that our proposed algorithm has achieved a much better performance than other deterministic algorithms.
 Index Terms - Graph Matching and Genetic Algorithm
Comparisons of Network Coding with Basic Routing Techniques
Pages: 7-17 (11) | [Full Text] PDF (423K)
Naushad M. PEYRYE and Mussawir A. HOSANY
Infosys Limited, Mauritius
Department of Electrical and Electronic Engineering, Faculty of Engineering, University of Mauritius, Mauritius

Abstract -
The Internet of today is based on a store-and-forward mechanism for packet transfer where routers have the ability of only forwarding packets to most appropriate links without modifying the data in the packets. However with the advent of Network Coding, data are encoded and decoded by routers before packets are forwarded, in view to reduce the bandwidth and thus to improve throughput. This paper presents the network coding principle as practically conceivable for routing by comparing it with basic static routing techniques through simulation. The result shows that network coding performance varies with network size and density. We also investigated a dynamic version of selective flooding and selective network coding through Routing State Information Ageing (RSIA) and the result shows some improvements to Flooding and inherently, but to some lower extent, to Network Coding.
 Index Terms - Network Coding, Routing, Flooding, Shortest Path and RSIA
New Approach for UML Based Modeling of Relational Databases
Pages: 18-23 (6) | [Full Text] PDF (516K)
Muhammad Rashid Naeem, Weihua Zhu and Adeel Akbar Memon
Software Engineering College, Chongqing University, Chongqing, P.R. China

Abstract -
In this paper, we propose new approach for UML based modeling of relational databases. Furthermore, we use UML Class Diagram and Sequence Diagram to model book data archive and SQL based operations on that data. However this approach is quite in general and can be used for other relational databases.
 Index Terms - Class Diagram, DML, Relational Databases, Sequence Diagram and UML
Security Issues in Web Services
Pages: 24-27 (4) | [Full Text] PDF (158K)
Hirrah Siddique and Syed Shah Muhammad
Department of Computer Science and Engineering, University of Engineering and Technology, Lahore-Pakistan

Abstract -
Computing based on Web services is currently very much popular and demanded in the software industry providing a variety of services that are required by various organizations and users. Due to the vastness and flexibility of the architecture the web services have to face a great number of threats and attacks and overall security issues. A secured architecture is very critical for the web services due to the dynamic nature of the system. Most of the work had done by researchers on the security issues for web services. In this paper, we shall review the security issues stated in the past and solutions of those issues, and then we shall present our findings on the security challenges that are present to the web services. MPMSD (Multi-Part Multi-Signature Document) is an emerging field of study and played an important role in the security of web services. In the paper, we will also discuss the MPMSD (Multi-Part Multi- Signature Document) and also identify the future challenges for the web services security.
 Index Terms - Web Services, SOAP, WSDL, XML, UDDI, SOC, MPMSD and Digital Signature
Image Resolution Enhanncement Using Sparse-Neighbor Embedding
Pages: 28-32 (5) | [Full Text] PDF (467K)
Priya B. and Gopakumar G.
Computer Science Department, College of Engineering, Chengannur

Abstract -
Obtaining a High-Resolution (HR) image of a scene from a single or multiple Low-Resolution (LR) images is known as image Super-Resolution (SR). A lot of applications like Computer Vision, Medical Imaging, Pattern Recognition, Satellite imaging, a robust performance can be achieved if the Low-Resolution input is super-resolved to a high-resolution image. Generally Neighbor-Embedding (NE) based super-resolution, first a neighbour search is performed using Euclidean distance after that optimal weights are determined using constrained-least square problem. In this paper a sparse-neighbor selection process is used for super-resolution reconstruction. Histogram of oriented Gradients (HoG) method is used to extract the features of Bicubically interpolated (BI) LR image patches. Then the whole training data set is partitioned into set of subsets by clustering the histogram of oriented gradients (HoG) using fuzzy C-means method. A Sparse neighbor selection algorithm (SPNS) is used to search by incorporating Robust-SL0 algorithm and classification-regression based super-resolution criterion. SPNE algorithm simultaneously searches the neighbors and estimation of weights is done.
 Index Terms - HoG, IBP and Super Resolution