Claim Missing Document
Check
Articles

Found 2 Documents
Search

Pengembangan Arsitektur Federated Learning untuk Privacy-Preserving Analytics pada Sistem IoT Skala Besar Abdurrohman Abdurrohman; Wemby Juniarrochman
Asian Journal of Multidisciplinary Research Vol. 3 No. 1 (2026): Asian Journal of Multidisciplinary Research
Publisher : Jujurnal Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59613/tdf9mr96

Abstract

Penelitian ini mengembangkan arsitektur Federated Learning (FL) hierarkis tiga tingkat untuk mengatasi berbagai tantangan pokok dan signifikan terkait privasi, skalabilitas, serta heterogenitas data pada sistem Internet of Things (IoT) skala besar. Melalui pendekatan terintegrasi, arsitektur yang diusulkan mengintegrasikan perangkat edge, gateway cluster, dan server cloud, yang dilengkapi dengan mekanisme privacy-preserving bertingkat menggabungkan differential privacy, secure aggregation, dan audit berbasis blockchain. Algoritma agregasi adaptif (FedAdapt) dirancang khusus untuk menangani data non-IID dan serangan Byzantine secara robust. Hasil eksperimen pada tiga domain aplikasi utama menunjukkan dengan jelas bahwa arsitektur mencapai akurasi rata-rata 92,8% serta yang menekan overhead komunikasi hingga 73,4% dibandingkan pendekatan standar. Sistem ini juga terbukti sangat tangguh terhadap serangan hingga 30% klien jahat, dengan tingkat keberhasilan serangan inferensi ditekan dari 72,3% menjadi 8,1%. Secara keseluruhan, arsitektur yang dikembangkan menawarkan keseimbangan optimal antara akurasi, efisiensi, dan privasi, menjadikannya solusi praktis dan skalabel untuk implementasi analitik data cerdas pada ekosistem IoT masa depan.
Development of a Smart Mobile Application for Student Time Management Abdurrohman Abdurrohman; Wemby Juniarrochman
Journal of Technology and Engineering Vol 4 No 1 (2026): Journal of Technology and Engineering
Publisher : Yayasan Banu Haji Samsudin

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59613/journaloftechnologyandengineering.v4i1.452

Abstract

This study explores the development of a smart mobile application for student time management through a library research approach. The purpose of the study is to identify the conceptual foundations, functional requirements, and development opportunities for a mobile application that can support students in managing their academic time more effectively. The reviewed literature shows that students frequently struggle with task organization, prioritization, and maintaining consistent study routines. The findings indicate that a smart mobile application should integrate scheduling, reminders, progress tracking, personalized suggestions, and user-friendly interfaces to enhance academic productivity. The literature also emphasizes the importance of behavioral support and contextual relevance in improving application effectiveness. Based on the synthesis of sources, the study concludes that a smart mobile application has strong potential to serve as an adaptive digital assistant for students. Although this study is limited to conceptual analysis, it provides a practical foundation for future prototype design and empirical testing.