Volume 5, Issue 8, August 2014

Distributed Memory and Shared Distributed Memory Architecture for Implementing Local Sequences Alignment: A Survey
Pages: 1-8 (8) | [Full Text] PDF (431K)
Manhal Elfadil Eltayeeb Elnour, Muhammad Shafie Abd Latif and Ismail Fauzi Isnin
Department of Computer Science, Faculty of Computing, Universiti Teknologi Malaysia, Johor, Malaysia

Abstract -
Recently, researchers have shown an increase interest in Sequences Alignment Problem (SAP), in which the properties of new DNA and Protein sequences are detected by comparing them with well-known reference sequences defined in genetic databanks. Considerable amounts of literature have been published on SAP. However, less attention has been paid to the revision and classification of current techniques in SAP. Thus, the purpose of this paper is to review recent trends in implementing local SAP in a parallel architecture. The classification is based on Distributed Memory and Shared Distributed Memory architectures for different Dynamic Programing and Smith-Waterman based algorithms. The literature are studying and evaluating in order to highlight their strength and weaknesses.
 Index Terms - Sequences Alignment, Shared Distributed Memory, DNA, Protein and Parallel Computing
Comparative Analysis of TDMA Scheduling Algorithms in Wireless Sensor Networks
Pages: 9-15 (7) | [Full Text] PDF (753K)
Palikhel Laxman and Prajapati Rajeev
Thapathali Campus, IOE Tribhuvan University, Kathmandu, Nepal
Department of Computer and Electronics Engineering, Kantipur Engineering College, Tribhuvan University, Kathmandu, Nepal

Abstract -
Energy conservation is a major issue in Wireless Sensor Network (WSN). In order to obtain energy conservation, Time Division Multiple Access (TDMA) has been discussed as one of the potential solution. Many researchers proposed TDMA as a Media Access Control (MAC) in order to conserve energy. The main advantage to using TDMA MAC is avoidance of collision of data packets during transmission and the added facility to use sleep modes. The use of sleep mode enables switching off the radio antennas thus reducing the energy conservation. Prior to usage of TDMA MAC, scheduling of the sensor nodes, i.e. providing time slot to the sensor nodes must be performed. Efficient scheduling of transmitting time slot in a TDMA is important for low power WSN. In this work, two decentralized scheduling algorithms Distributed Randomized TDMA (DRAND) and Deterministic Distributed TDMA (DDTDMA) are compared. In these algorithms, flowing the messages among the sensor nodes the scheduling is performed by assigning transmitting time slot to each node. So, their efficiency is analyzed based on schedule length, message complexity and convergence time to obtain scheduling. It was found that DDTDMA is an efficient algorithm than DRAND in terms of schedule length, message complexity and convergence time.
 Index Terms - Wireless Sensor Network, Time Division Multiple Access, Distributed Randomized TDMA and Deterministic Distributed TDMA
Traffic Engineering Based Load Balancing in LTE-A Heterogeneous Network
Pages: 16-25 (10) | [Full Text] PDF (1000K)
Engr. Rashid Farid and Junaid Arshad
Department of Computer Science and Engineering, University of Engineering and Technology, Lahore, Pakistan

Abstract -
LTE-A is the latest cellular network technology. LTE-A is the use of heterogeneous networks (HetNets) which contains combinations of different cells such as of a macro cells and low power nodes (LPNs). Heterogeneous Network is used to increase capacity of the system and as well as to increase the demand for data capacity, especially in hotspot areas where there is a high density of users. Load balancing is main issue in urban area, heterogeneous network is used for cell splitting gains and ensure for user experiences. Cell range extension (CRE) is a technique that can be used to achieve load balancing in heterogeneous network. By CRE offset is added to LPNs in cell selection, which expand the range of LPNs and offload many users from macro-cells to LPNs. CRE is used for uniform offsets. The result of uniform offset is not optimal in the load balancing of heterogeneous network. In this thesis I use the Heuristic load balancing algorithm which is designed for assigning cell specific offset LPNs. Heuristic methods are used to speed up the process of finding a suitable solution. A heuristic algorithm is used for good quality of load balancing which is close to optimal solution. By apply the concept of cell load coupling, Range Optimization framework algorithm is used for cell specific offset. In this research work we will implement our algorithm show results by using MATLAB based Vienna LTE simulator to check and compare results in different scenario using different offsets values.
 Index Terms - LTE, LTE-A, Heterogeneous Network (HetNets), LPN and CRE
The State of the Art Handwritten Recognition of Arabic Script Using Simplified Fuzzy ARTMAP and Hidden Markov Models
Pages: 26-32 (7) | [Full Text] PDF (626K)
Yusuf Perwej, Shaikh Abdul Hannan and Nikhat Akhtar
Department of Computer Science & Engg., Al Baha University, Al Baha, Kingdom of Saudi Arabia (KSA)
Department of Computer Science & Engg., Integral University, Lucknow, India

Abstract -
In this paper, we present recognition of handwritten characters of Arabic script. Arabic is now the 6th most spoken language in the world and is spoken by more than 200 million people worldwide. The 7th Century A.D., Arabic started to spread to the Middle East as many people started to convert to Islam. During this time of religious conversions, Arabic replaced many South Arabian languages, most of which are no longer commonly spoken or understood languages. The challenges in Arabic handwritten character recognition wholly lie in the variation and disfigurement of Arabic handwritten characters, since different Arabic people may use a different style of handwriting, and direction to draw the same shape of the characters of their known Arabic script. Though various new propensity and technologies come out in these days, still handwriting is playing an important role. To recognize Arabic handwritten data there are different strategies like Simplified Fuzzy ARTMAP and Hidden Markov Models (HMM). In this paper we are using Simplified Fuzzy ARTMAP, which is an updated version of Predictive Adaptive Resonance Theory. It also has a capacity to adjust clusters, as per the requirements Arabic script, which is remunerative to mitigate noise. We have tested our method on Arabic scripts and we have obtained encouraging results from our proposed technique.
 Index Terms - Hidden Markov Model (HMM), Arabic Script, handwriting, Fuzzy ARTMAP, Recognition and Feature Extractor
Situation Modeling Using Sentence Correlation and Importance
Pages: 33-38 (6) | [Full Text] PDF (218K)
Shahzeb Patel, Shreyas Talele, Abhijay Patne, Radhika Simant and Saurabh Karwa
College of Engineering, Pune–411005

Abstract -
Biomedical abstracts are abundant with information on medical entities. Large number of these abstracts are easily accessible and are used in the field of medical research. Extracting useful information out of these abstracts is a field of major interest among text mining community. Biomedical corpus form a major dataset of research in this area. This paper aims at exploring the development of a Situation Model (SM) on a set of such medical abstracts. It aims at finding, gathering, aggregating and analyzing information from these abstracts which is of utmost importance to the user, made readily available so that the user can see what the abstract aims at establishing. The work also tracks the relations within the text, protein interactions (PPI) to be precise which helps in developing a separate model of interest for the user. This paper details the steps to transform textual resources to a structured SM which covers integrated and exile representation of the available abstracts.
 Index Terms - Situation Modeling, Coherence, WordNet, Sentence Similarity and Word Importance Value
Challenges of Implementing and Developing E-Government: A Case Study of the Local Government System in Ghana
Pages: 39-48 (10) | [Full Text] PDF (304K)
Wisdom Kwawu Torgby and Nana Yaw Asabere
Department of Computer Science, School of Applied Sciences and Arts, Accra Polytechnic, Accra, Ghana

Abstract -
Although Electronic Government (E-Government) is a global phenomenon, the simple transfer of Information and Communication Technology (ICT) solutions and related organizational concepts from developed to developing countries does not seem very appropriate or feasible. With the current proliferation of ICT, there is no doubt that E-Government has the potential impact of reducing administrative and development problems in a developing or developed country. Nevertheless, it is evident that compared to developed countries, additional efforts are necessary when implementing E-Government in developing countries of which Ghana is no exception. In comparison to developed countries, the different cultural, institutional and wider administrative contexts of developing countries must be considered to circumvent unpremeditated effects. Using the local government system in Ghana and relevant literature, this article outlines and addresses the cultural and different institutional contexts which must be taken into account when implementing E-Government. The paper further discusses some key research challenges and open issues involving E-Government implementation in Ghana.
 Index Terms - E-Government, Ghana, ICT, Implementation and Local Government System
Efficient Load Balancing using VM Migration by QEMU-KVM
Pages: 49-53 (5) | Full Text] PDF (232K)
Sharang Telkikar, Shreyas Talele, Siddharth Vanarse and Amit Joshi
Department of Computer and IT Engineering, College of Engineering, Pune-411005

Abstract -
Load balancing is a computer networking method which is used to distribute the load across computer clusters or multiple computers, central processing units to achieve optimal resource utilization, maximize throughput, minimize response time and prevent overload. There are various methods to balance the load. This paper discusses the design and implementation of a policy engine to dynamically balance the load over a network, using live migration feature of KVM. The goal is to provide a provisioning monitor that can dynamically make decisions about migration of heavily/lightly loaded virtual machines.
 Index Terms - Load Balancing, Virtual Machine, Hypervisor and QEMU-KVM