http://repository.mut.ac.ke:8080/xmlui/handle/123456789/57 2026-04-07T22:04:46Z 2026-04-07T22:04:46Z Muthii, James Mburu Wanjau, Stephen K. Njenga, Stephen http://repository.mut.ac.ke:8080/xmlui/handle/123456789/6911 2026-04-01T08:38:07Z 2026-01-01T00:00:00Z

Scaffold Extensions for Client Drift Mitigation in Federated Learning: A Synthesis of Approaches, Limitations, and Future Directions Muthii, James Mburu; Wanjau, Stephen K.; Njenga, Stephen Client drift arising from non-independent and identically distributed (non-IID) data across participating clients remains one of the most critical obstacles to effective Federated Learning. The Scaffold algorithm, which introduces control variates to correct local gradient updates, has emerged as one of the most prominent variance reduction methods for mitigating this drift. Although numerous extensions to Scaffold have been proposed, no systematic review has exclusively examined the Scaffold algorithm and the control variate mechanism for client drift mitigation, leaving the research community without a consolidated understanding of how Scaffold has been extended, what limitations persist, and which characteristics remain underexplored. This study addresses that gap through a systematic literature review guided by PRISMA 2020 guidelines. Seven electronic databases were searched for publications from 2016 to 2026, yielding 1,847 records, from which 33 studies were included after duplicate removal, screening, and full-text eligibility assessment based on criteria requiring each study to address Scaffold or control variates for client drift in FL and cover at least two performance metrics. Data were synthesized thematically using frequency counts and tabular summaries. The review reveals nine distinct extension approaches: variance reduction via gradient estimation techniques was the most prevalent (11 studies, 34%), followed by integration with advanced optimization algorithms (8 studies, 25%), together accounting for 59% of the reviewed work. Twelve Scaffold characteristics were targeted for extension, with variance reduction the most commonly modified (37%, rising to 50% with combined categories), while communication mechanism, privacy budget allocation, and similarity-based approaches remained significantly underexplored. Recurring limitations across all approaches included communication and computational overhead, hyperparameter sensitivity, restrictive theoretical assumptions, performance degradation under extreme data heterogeneity, and limited large-scale empirical validation. A notable finding is that similarity-based approaches for client drift mitigation are largely absent from the literature, with only one study employing a similarity measure. The review, therefore, recommends future investigation of similarity-based methods as adaptive control variates within the Scaffold protocol, alongside prioritization of communication-efficient, privacy-preserving designs validated at scale. This research was self-sponsored with no external funding.

2026-01-01T00:00:00Z Muchori, Juliet Gathoni Kamau, Gabriel Ndung’u Ndung’u, Rachael http://repository.mut.ac.ke:8080/xmlui/handle/123456789/6874 2026-03-26T09:25:30Z 2025-01-01T00:00:00Z

System Log Parameter Attributes for Predicting Software Failures: Systematic Literature Review Muchori, Juliet Gathoni; Kamau, Gabriel Ndung’u; Ndung’u, Rachael The early prediction of software failure is important in the field of software engineering since it leads to the development of better quality software, along with a reduction in maintenance cost and effort. However, even though there is growing interest in early prediction of software failure, the existing literature shows some gaps. While many studies are quite reliant on static code metrics or test case execution data, they tend to miss out on vital dynamic and contextual information which can be obtained by analyzing software system logs. Log data is regularly created by computing systems during their runtime and contains rich information including event sequences, timestamps, error messages, and system states that can potentially being utilized in the identification of anomalies and predictions of failures on real-time. The objective of this work is to categorize the existing literature on the use of system logs for predicting software, through systematic literature review, with the help of the guidelines from Barbara Kitchenham. The review categorizes system logs into four primary parameters: resource/hardware logs, workload/performance logs, network logs, and security logs. It also highlights the machine learning models. The findings reveal that log attributes such as CPU usage, memory utilization, disk space, transaction processing, and network errors are consistently identified as key predictors of software failure. This finding aligns with expert opinions, demonstrating strong agreement on the relevance of these attributes for predicting software failure. This study contributes to the growing body of knowledge on software failure prediction, emphasizing the importance of integrating machine learning with systematic log monitoring to enhance proactive system failure management. Future work should focus on developing real-time monitoring tools that leverage machine learning models to automate failure detection and prediction across various system components.

2025-01-01T00:00:00Z Wambua, Anthony Wambua Kamau, Gabriel Ndung’u http://repository.mut.ac.ke:8080/xmlui/handle/123456789/6873 2026-03-26T09:18:35Z 2023-01-01T00:00:00Z

Security-aware Mobile Application Development Lifecycle (sMADLC) Wambua, Anthony Wambua; Kamau, Gabriel Ndung’u With the high mobile phone penetration and subsequent significant usage of mobile phone applications, mobile users have become prime targets of hackers. Secure Software Development (SSD) advocates incorporating security aspects at the initial stages of software development. This study proposes a novel Mobile Application Development Lifecycle by reviewing SSD concepts and incorporating these concepts into MADLC- a mobile-focused software development lifecycle to create a security-aware Mobile Application Development Lifecycle (sMADLC). The proposed development lifecycle, sMADLC, can potentially help mobile application developers create secure software that can withstand hacker aggression and assure mobile application users of the confidentiality, integrity and availability of their data and systems.

2023-01-01T00:00:00Z Walumbe, Denis Wapukha Kamau, Gabriel Ndung’u Njuki, Jane Wanjiru http://repository.mut.ac.ke:8080/xmlui/handle/123456789/6872 2026-03-26T09:13:59Z 2025-01-01T00:00:00Z

Lightweight Proof-of-Stake Models for PrivacyPreserving Telemedicine Systems: A Systematic Review Walumbe, Denis Wapukha; Kamau, Gabriel Ndung’u; Njuki, Jane Wanjiru Proof of Stake (PoS) models are energy-efficient and require limited computational power. These features are critical in telemedicine environments, where resource-constrained devices must handle sensitive data securely. The growing need for auditable and privacy-preserving data storage in telemedicine underscores the importance of PoS models optimized for lightweight devices while complying with strict regulatory requirements, such as the Health Insurance Portability and Accountability Act (HIPAA).This study was guided by two research questions: (i) Which PoS models are lightweight and suitable for telemedicine? and (ii) What features make lightweight PoS models effective for privacy and efficiency in telemedicine? To address these questions, a systematic literature review (SLR) guided by the PICOC framework was conducted to investigate lightweight PoS models that can enhance privacy in telemedicine systems. Out of 2,394 papers studies screened, 55 were included in the analysis. The findings identified Algorand, Ouroboros Praos, Tendermint, Nxt, and Casper CBC as promising candidates. Key enabling features included lightweight voting mechanisms, such as Byzantine Agreement protocols and Verifiable Random Functions, as well as cryptographic techniques like symmetric encryption and multiparty computation. Performance metrics evaluated included latency, throughput, energy efficiency, and battery consumption, with Grey Relational Analysis ranking Algorand highest due to its low latency, high throughput, and minimal energy consumption

2025-01-01T00:00:00Z