SYSTEM INTEGRATION AND BACKEND IMPLEMENTATION FOR ENHANCHING SLA MONITORING IN UNDERDEVELOPED REGIONS OF INDONESIA USING LSTM­ BASED PREDICTION

• Main Author: Royyan, Aulia
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/87885
• Institution: Institut Teknologi Bandung

This research explores the development of a web application system for prediction based on the LSTM (Long Short-Term Memory) algorithm in the telecommunications industry, specifically in the 3T regions (Underdeveloped, Frontier, and Outermost areas) oflndonesia. The 3T regions face unique challenges in delivering telecommunications services, such as achieving equitable services and also ensuring reliable services. The government has made numerous programs to provide telecommunications services, particularly intemet access, in these regions. And in the practice, the SLA (Service-Level Agreement) and its monitoring processes become crucial and highly relevant aspects. Therefore, the authors identify an opportunity to utilize web application systems and machine leaming to enhance SLA monitoring systems, thereby improving the delivery of reliable and equitable telecommunications services in the 3T regions across Indonesia. This paper focuses on system integration in the development of web applications and the LSTM machine leaming model to assist in SLA monitoring processes. The LSTM model is employed due to its superior ability to process and analyze large­scale time-series data. This model predicts and identifies pattems and anomalies in network performance during monitoring processes, which serve as supporting parameters to understand the provision of intemet services and SLA at any given time. Consequently, it aids in preparation and decision-making processes for service provision programs. Furthermore, system integration and web application development ensure a coherent and logically straightforward system for users. Web application development facilitates user-friendly accessibility. Coupled with cloud platform utilization, this system supports easy and flexible scalability, as it minimizes the need for extensive computing resources for the user. Backend development is carried out using the Flask framework with Python programming language to comprehensively integrate the system logic. Frontend development is conducted using React with JavaScript programming language to present system functionalities in a user-friendly manner. System integration with the cloud platform using AWS (Amazon Web Services) enables easy access and scalability of the system services, as well as lightweight computational demands for users. The conclusion of this research supports the potential use of integrating web application systems and LSTM machine learning models to facilitate SLA monitoring and management processes in the telecommunications sector, especially in the 3T regions of Indonesia. The test results indicate significant ease in data processing and identifying issues in the performance of intemet service provision from monitoring. Additionally, the system's lightweight and user-friendly features help ensure accessibility for users, particularly in 3T regions that may have limited resources. 1n the broader scheme, this research has significant implications for digital rights and telecommunications services equalization across Indonesia. It contributes to improving telecommunications infrastructure in the 3T regions of Indonesia, indirectly playing a role in narrowing socioeconomic disparities and advancing human resource development in Indonesia in general.