Communities in ResearchSpace

Welcome to ResearchSpace, the institutional repository of the CSIR. ResearchSpace is an open access electronic archive collecting, preserving and distributing scholarly digital materials created by the CSIR.

Most Viewed Items
Views

Recent Submissions

Item
Reliability evaluation of RSrSF-LoRa and LoRaWAN for dense industrial IoT networks in a smelter environment
(2026-08) Musonda, SK; Ndiaye, M; Simate, Z; Libati, HM; Abu-Mahfouz, Adnan MI
Industrial Internet of Things (IIoT) systems are increasingly deployed in heavy industrial facilities to support real-time monitoring, automation, and safety-critical operations. Long-range low-power communication technologies such as LoRaWAN are widely used for large-scale sensor connectivity due to their long communication range and energy efficiency. However, dense industrial deployments may experience reduced communication reliability due to network congestion, packet collisions, and challenging propagation conditions caused by metallic infrastructure and electromagnetic interference. To address these limitations, this paper proposes RSrSF-LoRa, an integrated implementation of the reserved spreading factor (rSF) mechanism within the RS-LoRa framework (lightweight scheduling), designed to improve reliability for critical traffic. The study presents a comparative performance evaluation of RSrSF-LoRa and LoRaWAN in a smelter IIoT deployment, assessing packet delivery ratio (PDR), throughput, fairness, energy consumption and scalability under varying node densities and gateway configurations. Single-gateway scenarios with 100, 500, and 1000 nodes, and seven-gateway scenarios with 1000 and 2100 nodes, are evaluated as communication distances increase. The results indicate that while throughput and fairness remain comparable across approaches, RSrSF-LoRa improves alarm message reliability in dense single-gateway deployments, extending the acceptable PDR from 700 m to 900 m. Alarm nodes in RSrSF-LoRa consume slightly more energy due to reserved transmission, but overall energy consumption remains comparable. These findings provide design insights for reliable and energy-aware industrial IoT networks in smelter environments.
Item
Discriminant analysis-guided multi-split segmentation for expectation maximization initialization in timeseries data
(2026-09) Onumanyi, Adeiza J; Abu-Mahfouz, Adnan MI; Kurien, Anish
In this paper, a discriminant analysis–guided multi-split segmentation approach referred to as K-between class segmentation (K-BCS) is presented, which is developed as an effective initialization strategy for the expectation–maximization (EM) algorithm in time-series data. The proposed approach is aimed at improving parameter estimation and convergence stability in the EM algorithm. Although EM is widely used for mixture modeling, its performance is highly sensitive to its initialization estimates, which often results in slow convergence, convergence to suboptimal local maxima, and unstable likelihood behavior. To address these limitations, K-BCSEM leverages a proposed between-class measure to identify intrinsic data structures and generate informed initial parameter estimates. Unlike random or centroid-based methods, the proposed approach is fully datadriven and captures separation in both location and dispersion, thereby providing more reliable starting points for the EM algorithm. The proposed approach was evaluated via extensive Monte Carlo experiments across both synthetic and real-world datasets. The results obtained show that the K-BCS-EM approach consistently outperformed conventional initialization strategies, such as random sampling, random partitioning, jittered sampling, and k-means-based methods particularly in terms of their likelihood convergence, mean and variance RMSE performance. The K-BCS-EM demonstrates strong robustness in challenging scenarios such as in heavy-tailed distributions and highly overlapping mixtures while maintaining scalability as the number of components increases. Although the proposed approach introduces a modest computational overhead due to its preprocessing stage, the significant improvements in estimation accuracy, convergence stability, and robustness highlight K-BCS-EM as an effective initialization method for EM-based mixture modeling in time series data.
Item
Technology Predictions 2026
(2026-04) Ebert, C; El Hajj, I; Frachtenberg, E; Lysko, Albert A; Milojicic, D; Saint Nom, R; Sinha, S; Toro, J
For the past 16 years, every January, the IEEE Computer Society has identified technologies most likely to succeed in the year ahead. In recent years, this effort has expanded and formalized beyond traditional computing, covering areas, such as power and energy, space, health technology, and more. At the same time, we have strengthened the team’s global representation and gender diversity to ensure these predictions reflect the full breadth of innovation and impact worldwide.
Item
Exploring social media influencers promotion of online gambling among South African youth
(2026-05) Baloyi, Errol; Siphambili, Nokuthaba; Latakgomo, Molebogeng; Mukondeleli, Elekanyani
The COVID-19 pandemic has accelerated the digitalization of daily life, shifting work, education, and social interaction into the online environment. Within this context, social media has become a powerful platform for marketing and behavioural influence, reshaping how young South Africans engage with online gambling. This study examines the role of social media influencers in promoting and legitimizing online gambling among South African youth through a qualitative content analysis of posts on Instagram, TikTok, and Facebook. The findings reveal that gambling-related content is frequently embedded within lifestyle, entertainment, and aspirational narratives that portray gambling as a routine, glamorous, and desirable activity. Such portrayals contribute to the normalization of gambling, blurring the line between leisure and risk-taking behaviour. The findings further highlight the growing social and psychological consequences of this trend, including financial vulnerability, addiction risks, and cases in which some South African higher education students reportedly use their National Student Financial Aid Scheme (NSFAS) allowances to fund online gambling activities. Moreover, the study identifies significant regulatory gaps between current social media marketing practices and existing South African gambling legislation, exposing weaknesses in the oversight of influencer advertising. These findings underscore the urgent need for stronger regulatory frameworks and ethical standards governing digital marketing and gambling promotion. The paper concludes with recommendations for policymakers, regulators, and digital platforms to strengthen advertising controls, enhance consumer protection, and promote media literacy initiatives aimed at mitigating the harmful effects of influencer-driven gambling promotion among South African youth
Item
Time-Interleaved noise jamming
(2022-12) Claasen, Gert; Du Plessis, Warren P
Modern jamming systems are faced with the reality that multiple threat radars will be encountered simultaneously, and potentially, all simultaneously encountered threat radars will need to be countered to prevent detection. Modern jammer systems are capable of extremely rapid reconfiguration, allowing them to counter multiple simultaneous threats in a time-interleaved manner. However, simultaneously encountered threats cause jamming pulses to coincide in time, leading to the problem of deciding which of the threats to jam when such coincidences occur. In this article, an approach to determining the relative priority of jammer pulses is proposed and evaluated to show that time-interleaved jamming is a viable approach to effectively countering multiple simultaneous threats.