Rancang Bangun Modul Otentikasi Dan Otorisasi Berbasis Rekognisi Wajah Dengan Arsitektur Microservices Dan API Gateway Pada LMS Bahasa Inggris

Sanjaya, I Gusti Agung Ngurah Adhi (2025) Rancang Bangun Modul Otentikasi Dan Otorisasi Berbasis Rekognisi Wajah Dengan Arsitektur Microservices Dan API Gateway Pada LMS Bahasa Inggris. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5025211056-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (11MB) | Request a copy

Abstract

Kemajuan teknologi informasi membuat kebutuhan akan sistem keamanan digital yang andal semakin penting dalam aplikasi pembelajaran daring (LMS). Pada penelitian sebelumnya menerapkan sistem otentikasi email dan password yang masih kurang aman dan rawan disalahgunakan. Pada penelitian sebelumnya sudah menerapkan arsitektur microservices namun belum diterapkan secara modular, sehingga membatasi skalabilitas dan fleksibilitas sistem. Penelitian ini mengembangkan sistem otentikasi dan otorisasi berbasis microservices dengan integrasi Face Recognition menggunakan DeepFace dan API Gateway sebagai penghubung antar layanan. Sistem terdiri dari empat layanan utama: User Management, Role Management, Authentication & Security, dan API Gateway, yang dibangun dengan framework Flask. Otentikasi biometrik dilakukan dengan metode face verification dan skema multireference, serta dilengkapi pengelolaan token khusus untuk akses fitur krusial. Hasil pengujian fungsional menunjukkan bahwa sistem berfungsi baik untuk tiga peran utama: Guru, Siswa, dan Admin. Layanan Role Management dan User Management telah berjalan secara independen, sedangkan Authentication & Security masih memiliki ketergantungan data. Dari sisi performa, pengujian non-fungsional mencatat hasil stabil pada fitur-fitur setiap layanan salah satunya fitur get profile mendapatkan nilai Response time 154 milidetik, Throughput 6,1 permintaan/detik, dan Error rate 0%. Namun, fitur login-face biometrik menunjukkan penurunan performa pada server yang hanya berbasis CPU tanpa GPU dengan Response time sebesar 41036 milidetik, Throughput sebesar 0,8 permintaan/detik, dan Error rate sebesar 84%. Dengan pendekatan ini, sistem menawarkan keamanan yang lebih baik, modularitas layanan, serta fleksibilitas otentikasi yang cocok untuk pengembangan LMS yang lebih cerdas dan aman selanjutnya.
===================================================================================================================================
The advancement of information technology has increased the need for reliable digital security systems, particularly in online learning applications (LMS). Previous studies implemented email and password-based authentication, which remain vulnerable to misuse and security threats. Although microservices architecture had been applied, it was not yet implemented modularly, limiting the system’s scalability and flexibility. This study develops an authentication and authorization system based on microservices, integrating Face Recognition using DeepFace and an API Gateway to connect between services. The system consists of four main services: User Management, Role Management, Authentication & Security, and API Gateway, all built using the Flask framework. Biometric authentication is performed through face verification using a multi-reference scheme, complemented by a special token mechanism for accessing crucial features. Functional testing shows that the system works effectively for the three main user roles: Teacher, Student, and Admin. The Role Management and User Management services operate independently, while Authentication & Security still has data dependency. From a performance standpoint, non-functional testing reports stable results across service features — for instance, the get profile feature achieved a Response time of 154ms, Throughput of 6.1 req/s, and Error rate of 0%. However, the biometric login-face feature exhibited decreased performance when run on CPU-only servers without GPU support, recording a Response time of 41036ms, Throughput of 0.8 req/s, and Error rate of 84%. This approach offers improved security, modular service design, and flexible authentication, making it suitable for the future development of smarter and more secure LMS platforms.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	API Gateway, DeepFace, Flask, Microservices, Otentikasi, Otorisasi, Pengenalan Wajah, Authentication, Authorization, Face Recognition
Subjects:	T Technology > T Technology (General) T Technology > T Technology (General) > T58.6 Management information systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	I Gusti Agung Ngurah Adhi Sanjaya
Date Deposited:	29 Jul 2025 05:45
Last Modified:	29 Jul 2025 05:45
URI:	http://repository.its.ac.id/id/eprint/123115

Actions (login required)

View Item