DUT Open Scholar:: Home

Open Scholar is the institutional repository of Durban University of Technology. As far as possible, it provides open access to the academic output of the University.

item types	authors	articles	theses

Some publishers do not permit the archiving of your article in institutional repositories if they have published it:

BASE is one of the world's most voluminous search engines especially for academic web resources:

https://www.base-search.net/

Communities in Open Scholar

Select a community to browse its collections.

Now showing 1 - 5 of 12

Recent Submissions

The relationship between physical activity and cardiorespiratory fitness among people living with Human Immunodeficiency Virus throughout the life span
(Ovid Technologies (Wolters Kluwer Health), 2019) Webel, Allison R.; Perazzo, Joseph; Phillips, J. Craig; Nokes, Kathleen M.; Rentrope, Cynthia; Schnall, Rebecca; Musanti, Rita; Adams Tufts, Kimberly; Sefcik, Elizabeth; Hamilton, Mary Jane; Portillo, Carmen; Chaiphibalsarisdi, Puangtip; Orton, Penelope; Davis, Liana; Rose, Carol Dawson
Background - People living with human immunodeficiency virus (PLHIV) are at an increased risk for developing cardiovascular disease (CVD). Physical activity and cardiorespiratory fitness in PLHIV are poorly understood.<h4>Objective</h4>The aims of this study were to describe physical activity and cardiorespiratory fitness by sex and age and to examine the association between physical activity and cardiorespiratory fitness in PLHIV, controlling for covariates. Methods Seven hundred two PLHIV participated in a cross-sectional study and completed validated measures of self-reported physical activity (7-day Physical Activity Recall) and cardiorespiratory fitness (6-minute walk test). Participants were recruited from 7 diverse sites in the United States and Thailand, and data were analyzed using descriptive statistics and multiple regression to examine the relationship between physical activity and cardiorespiratory fitness. Results On average, participants self-reported engaging in 115 minutes of, mostly light (75%), physical activity. Men reported twice the amount of physical activity as women (155 vs 73 minutes, P = .01). Participants' ability to achieve their predicted 6-minute walk test distances was similar between men (68%) and women (69%) (P > .01). For women, vigorous physical activity was associated with a 6.6% increase in cardiorespiratory fitness and being temporarily unemployed was associated with an 18% decline in cardiorespiratory fitness. Cardiorespiratory fitness increased with age (P < .01). Conclusion Weekly physical activity of people living with human immunodeficiency virus averaged 85 minutes of mostly light activity, well below the recommended 150 minutes of moderate activity. Vigorous physical activity was associated with improved cardiorespiratory fitness in women, but not men. Although PLHIV would benefit from interventions to increase physical activity, our data suggest a need to develop sex-specific physical activity strategies.
Investigation of path loss in a complex building environment using USRP and GNU radio
(Engineering and Technology Publishing, 2024-1-1) Itamakinde, Akinyele A; Ejidokun, Temitayo O; Adejumobi, Babatunde S; Shongwe, Thokozani; Adetiba, Emmanuel; Owolabi, Israel Esan
This paper presents the description, measurements, and analysis of an investigation into the path loss channel characteristics of a complex building environment using the College of Engineering building of Afe Babalola University, Ado-Ekiti, Nigeria as a case study. The study is conducted on the first floor of the building, which is compartmentalized into five scenarios. The measurements are performed in the distinctive line-of-sight and non-line-ofsight paths. Over 90% of the measurements taken are in the non-line-of-sight environment of the complex building. The wideband sliding correlator channel sounder measurement technique of 1.2 GHz centre frequency was performed using the National Instrument Universal Software Radio Peripheral (NI USRP) 2920/GNU (GNU’s Not Unix) radio testbed. Key statistical considerations associated with path loss are calculated and the path loss equation model for the non-line-of-sight and line-of-sight of the entire first floor of the complex building is obtained. The results of the investigation show that an increase in path loss is a function of distance. The calculated path loss exponents were used to rate the performance of the wireless communication channel. The values obtained were found to be in tandem with the existing path loss metrics. Also, the radio propagation along the enclosed passages and indoor environments experienced low path loss, while the radio propagation along the ornamental trees encountered large path loss. Consequently, the results of the linear regression model and the log-normal model equation of Nonline of Sight (NLOS) and Line of Sight (LOS) results are somehow correlated. These results are useful in predicting the path loss of the radio signal at any specific distance from the transmitter to the receiver in any environment similar to the one under investigation. Also, as a planning and coverage optimization tool for wireless communication designers.
A spectrum sensing and allocation model for primary user detection and interference mitigation in television whitespaces
(Science Publications, 2024-3-1) Notcker, Joachim; Adetiba, Emmanuel; Ronoh, Kennedy Kibet; Abayomi, Abdultaofeek; Akinola, Olubunmi Adewale
Television White Space (TVWS) emerges as an encouraging solution to address the challenge of a restricted wireless communication spectrum. It denotes the frequency range spanning from 54–790 MHz and researchers have increasingly explored its propagation characteristics in recent years. Nonetheless, a notable hindrance to its effective utilization lies in the interference between primary and secondary users, as well as interference among secondary users themselves. Approaches involving spectrum sensing and resource allocation have been extensively employed independently to tackle these issues, yet they have not been integrated or utilized in combination. Hence, in this study, we formulated an architectural model that combines spectrum sensing and allocation components. This integrated model aims to detect the presence of primary users while simultaneously minimizing interference among secondary users. The spectrum sensing component utilized an energy detection model to identify primary users, mitigating interference with secondary users. Meanwhile, the spectrum allocation component employed the Particle Swarm Optimization (PSO) algorithm to determine the optimal distribution of channels among secondary users. We implemented the architectural model in a simulated TVWS network using MATLAB R2020a. Its performance was then evaluated and compared with that of matched filter and Artificial Bee Colony (ABC) algorithms, which were utilized for spectrum sensing and allocation, respectively. Based on the simulation findings, when the Signal-to-Noise Ratio (SNR) was configured at -10 dB, the detection probability for the energy detection model reached 98.23%, surpassing the matched filter’s detection probability of 92.55%. With a false alarm probability of 0.51, the energy detection model exhibited a misdetection probability of 0.13%, outperforming the matched filter which had a higher misdetection probability of 2.61%. In scenarios with 10 channels and 100 secondary users, the particle swarm optimization algorithm attained a maximum throughput of 279.9 Mbps, slightly outperforming the artificial bee colony algorithm, which achieved 278.7 Mbps. In scenarios with 30 channels and 200 secondary users, the particle swarm optimization algorithm achieved throughputs of 1.575 Gbps, whereas the artificial bee colony algorithm achieved a comparable throughput of 1.571 Gbps. In the scenario where the number of channels was set to 50 and users to 300, the particle swarm optimization algorithm achieved a throughput of 3.879 Gbps, slightly surpassing the artificial bee colony algorithm, which achieved 3.864 Gbps. While the designed components consistently outperformed the matched filter and artificial bee colony algorithms across all cases, it’s important to note that the model faced limitations. Specifically, it was unable to detect more than one primary user or allocate spectrum for a new incoming secondary user.
Analysis and measurement of Tuberculin skin test induration using deep neural network
(International Association of Online Engineering (IAOE), 2024-9-13) Akinola, Olubunmi Adewale; Orimolade, Joseph Folorunsho; Afolabi, Akindele Segun; Shopeju, Habeeb Kehinde; Adetiba, Emmanuel; Adewale, Adeyinka Ajao
The World Health Organization (WHO) posited that tuberculosis (TB) is among the world’s ten greatest causes of mortality. Early case identification and timely treatment could minimize TB morbidity and death rates. This study adopts the UNets model for automatically detecting TB in subjects by using a deep neural network to assess the size of induration after tuberculin was injected into their hands. In order to do this, two neural network models were fine-tuned utilizing pre-learned weights from the 2012 ILSVRC ImageNet. Algorithms were developed to perform semantic segmentation of induration and compare it to that of a reference object of a known dimension. This was used to classify the status of the subject as either positive or negative. A series of experiments performed demonstrated that the optimal selection of neural network hyperparameters may provide a satisfactorily high F1 score of up to 0.977.
Education and the Fourth Industrial Revolution: eight years later
(Research India Publications, 2024-10-31) Kehdinga, George Fomunyam
This paper delves into the transformative impact of the Fourth Industrial Revolution (4IR) on various sectors, with a particular focus on education. Divided into three comprehensive sections, the first part explores the dynamics of 4IR, detailing the technological advancements and innovations that characterise this era, such as artificial intelligence, robotics, the Internet of Things (IoT), and big data analytics. The second section examines the application of these 4IR technologies in the educational landscape, highlighting how they are reshaping teaching methodologies, learning environments, and administrative processes. This includes an analysis of smart classrooms, personalised learning experiences, and the integration of digital tools in curriculum delivery. The third section envisions future possibilities, considering the potential long-term impacts of 4IR technologies on education and society at large. It discusses emerging trends, potential challenges, and opportunities for further innovation. The paper also addresses some arguments against the existence of an Industrial Revolution, questioning whether the changes we are witnessing constitute a true revolution or merely an evolution of existing technologies. The conclusion summaries the key discussions, emphasising the need for adaptive strategies to harness the full potential of 4IR in education while critically evaluating the concept of an industrial revolution itself. Through this exploration, the paper aims to provide a comprehensive understanding of how 4IR is revolutionising education and what the future may hold.