Volume: 12
Number: 1
June 2015

To see an article directly, click its Title. To see abstract, click its [Abstract] link.

• Software Defect Prediction Using Minimized Attributes
  Md. Habibur Rahman, Sadia Sharmin, Sheikh Muhammad Sarwar, Shah Mostafa Khaled, and Mohammad Shoyaib
  Abstract PDF
  Software Defect Prediction Using Minimized Attributes
  Md. Habibur Rahman, Sadia Sharmin, Sheikh Muhammad Sarwar, Shah Mostafa Khaled, and Mohammad Shoyaib

  Abstract

  Software quality estimation requires the identification of the number of defects that exist in a software. A software can be represented by a set of static code attributes and these attributes can be used to determine the defectiveness using simple statistical and machine learning tools. Among those attributes, all are not equally responsible for determining whether a software is defective or not. In this paper, we propose an attribute selection technique to select the most responsible attributes for building a defect prediction model. The model is experimented for both the within project and cross-project defect prediction using NASA Metric datasets and Relink. We observed a significant improvement in accuracy for both the within project and cross-project defect prediction, which proves the effectiveness of our proposed model.
• Hybrid Single Phase Buck AC-DC Converter with Improved Power Quality
  Golam Sarowar, Md Ashraful Hoque, and Mohammad Ali Choudhury
  Abstract PDF
  Hybrid Single Phase Buck AC-DC Converter with Improved Power Quality
  Golam Sarowar, Md Ashraful Hoque, and Mohammad Ali Choudhury

  Abstract

  A new topology of single-phase AC-DC converter using Buck conversion with high efficiency at extremely low duty cycle is proposed with low input current total harmonic distortion (THD) and high input power factor. Proposed double stage converter consists of single phase rectifier followed by a switched capacitor buck DC-DC converter. The input current THD is kept low and the input power factor is high with two-loop feedback control. The proposed schemes can be used in the application of new generation LED lighting.
• RARRA: Receiver-Assisted Robust Rate Adaptation in Wireless Networks
  Nafiul Rashid, Syed Sabir Salman-Al-Musawi, and Muhammad Mahbub Alam
  Abstract PDF
  RARRA: Receiver-Assisted Robust Rate Adaptation in Wireless Networks
  Nafiul Rashid, Syed Sabir Salman-Al-Musawi, and Muhammad Mahbub Alam

  Abstract

  The IEEE 802.11 wireless local area net- work (WLAN) standard, especially 802.11a remains the most popular way to exchange data over wireless links. The major requirement is to adapt to highly dynamic channel conditions with minimum overhead and ensure robustness and speed of transmission. How- ever, switching to the optimal transmission rate is a problem as 802.11 specification fails to specify via SNR measurement, the accurate channel condition at receiver. A further problem lies in calibrating such SNR values to the optimal rate. To this end we propose a novel Rate Adaption Scheme RARRA (Receiver Assisted Robust Rate Adaptation). Our key contributions include exploiting the more precise channel estimation of SNR- based Rate Adaptation coupled with estimating the channel condition at the receiver and finally sending this estimated information to the transmitter with minimum overhead. In other words we avoid RTS/CTS overhead to send the channel condition to the transmitter and use acknowledgment rates to serve‘ this purpose. Secondly, we differentiate the cause of frame loss as either due to channel error or collision using RTS/CTS in an adaptive fashion. This minimizes overhead but at the same time ensures that rate is not falsely changed due to frame loss caused by collision. RARRA exploits the best of SNR based approaches and provides channel condition at the receiver to the transmitter with minimum overhead aided by Adaptive RTS.
• CFD Analysis of Brick Klin Flue Gases and its Health Impact: A Case Study of Northern Part of Dhaka, Bangladesh
  Shahriar Shams, M Hamidur Rahman, and J. N. Sahu
  Abstract PDF
  CFD Analysis of Brick Klin Flue Gases and its Health Impact: A Case Study of Northern Part of Dhaka, Bangladesh
  Shahriar Shams, M Hamidur Rahman, and J. N. Sahu

  Abstract

  The northern part of Dhaka city in Bangladesh has contributed to air pollution due to a large number of brick kilns especially operating during the manufacturing season starting from October to March. The study looks at the distribution pattern of emission coming out from the chimneys of brick kilns through computational fluid dynamics (CFD) analysis and its possible impact on the surrounding environment. CFD analysis shows that when wind velocity is more than 3 m/s with increased exit velocity varying from 7.5 to 9 m/s the flow pattern of pollutants are more or less the same and surrounding buildings and vegetation are very likely to be affected. Whereas, if the wind velocity is less than 1 m/s with increased exit velocity varying from 6 to 17.5 m/s, the flow pattern of pollutants are rather dispersive and gradually more inclined towards its height of the chimney and very unlikely the surrounding buildings and areas are directly to be affected.
• Comparative Speed Study: A Way to Improve Road Safety Condition
  Shakil Mohammad Rifaat, Mosabbir Pasha, Mahmudul Hasan Shovon, Sajidur Rah-man Nafis, and Md. Kamruzzaman Haque Limon
  Abstract PDF
  Comparative Speed Study: A Way to Improve Road Safety Condition
  Shakil Mohammad Rifaat, Mosabbir Pasha, Mahmudul Hasan Shovon, Sajidur Rah-man Nafis, and Md. Kamruzzaman Haque Limon

  Abstract

  With the development of science and tech- nology people experience a large number of motor vehicles which is the main and easier way of travelling for people and carrying goods from one place to another. Roads, highways and streets are fundamental to trans- portation system and over 70% of passenger travel and much of goods mover is over the highways. However, at the same time we are sacrificing a large number of valuable human lives or being severely injured each day around the world. Road crashes are the major cause for it. It is a leading cause of deaths and serious injuries. Several studies justifies that speed is the major cause related to the risk of getting involved in a road accident. However, very few speed related studies done in the developing countries context, for example, Bangladesh. Specifically, the road safety authority of this country does not aware how speed data can be effectively used to enhance safety. This paper investigates the speed scenario of a particular roadway section of Mirpur Road as a pilot project. The speed data of various types of vehicles that use this particular roadway were taken for this study. To get a complete speed scenario of this road section data were taken on both weekdays and weekends in four different time periods that include morning peak (8.00a.m-9.00a.m), off-peak (11.00a.m- 12.00p.m), evening peak (6.00p.m-7.00p.m) and night time (9.00p.m-10.00p.m). Unpaired T- test analysis was performed considering 90% confidence interval for both equal and unequal variances. It is found that compared with other time periods vehicle speed varies in the morning-peak at weekdays and in the evening-peak at weekends. The speed of car, bus and truck significantly varies between two different days. Some measures that demand priority consideration in improving road safety are increasing road width, deploying traffic police in particular time periods, build speed bump, provide ad- equate streetlight, installing optimum number of speed limit sign etc.