SISTEM MONITORING DAN PREDIKSI DOWNTIME MODEM BERBASIS ENERGI SURYA MENGGUNAKAN SNMP (SIMPLE NETWORK MANAGEMENT PROTOCOL)

Geralda Livia Nugraha, - and Mahmudah Salwa Gianti, - and Dewi Indriati Hadi Putri, - (2025) SISTEM MONITORING DAN PREDIKSI DOWNTIME MODEM BERBASIS ENERGI SURYA MENGGUNAKAN SNMP (SIMPLE NETWORK MANAGEMENT PROTOCOL). S1 thesis, Universitas Pendidikan Indonesia.

Abstract

Keterbatasan pasokan listrik di wilayah 3T Indonesia menghambat operasional infrastruktur telekomunikasi, termasuk modem berbasis energi surya. Penelitian ini mengembangkan sistem monitoring dan prediksi downtime modem berbasis PV menggunakan protokol SNMP. Hardware terdiri dari Arduino Nano V3, sensor DHT22, BH1750, tegangan, HSTS016L, serta Orange Pi Zero 3. Data dikirim melalui RS485 ke Orange Pi, kemudian diteruskan via SNMP ke PRTG Network Monitor. Prediksi downtime dilakukan dengan algoritma Logistic Regression dan Random Forest menggunakan data historis sensor dengan pendekatan TimeSeriesSplit. Pengujian model menunjukkan Random Forest unggul dengan Downtime Recall 93,24% dan AUC 98,50%, dibandingkan Logistic Regression dengan Downtime Recall 58,11% dan AUC 91,30%. Visualisasi pada PRTG memudahkan deteksi anomali, sedangkan model prediktif mampu mengidentifikasi 93,24% kasus downtime untuk mencegah gangguan tak terduga. Integrasi ini terbukti efektif untuk predictive maintenance, meminimalkan downtime, dan menjaga kontinuitas layanan modem di wilayah terpencil, mendukung pemerataan akses telekomunikasi berbasis energi terbarukan. ----- Limited electricity supply in Indonesia's 3T regions hinders telecommunications infrastructure operations, including solar-powered modems. This research develops a PV-based modem monitoring and downtime prediction system using SNMP protocol. Hardware consists of Arduino Nano V3, DHT22, BH1750, voltage, HSTS016L sensors, and Orange Pi Zero 3. Data is transmitted via RS485 to Orange Pi, then forwarded via SNMP to PRTG Network Monitor. Downtime prediction is performed using Logistic Regression and Random Forest algorithms with historical sensor data using TimeSeriesSplit approach. Model testing shows Random Forest outperforms with Downtime Recall 93.24% and AUC 98.50%, compared to Logistic Regression with Downtime Recall 58.11% and AUC 91.30%. PRTG visualization facilitates anomaly detection, while predictive models can identify 93.24% of downtime cases to prevent unexpected disruptions. This integration proves effective for predictive maintenance, minimizing downtime, and maintaining modem service continuity in remote areas, supporting equitable access to renewable energy-based telecommunications.

	Text S_MKB_2100179_Title.pdf Download (514kB)
	Text S_MKB_2100179_Chapter1.pdf Download (292kB)
	Text S_MKB_2100179_Chapter2.pdf Restricted to Staf Perpustakaan Download (1MB)
	Text S_MKB_2100179_Chapter3.pdf Download (1MB)
	Text S_MKB_2100179_Chapter4.pdf Restricted to Staf Perpustakaan Download (2MB)
	Text S_MKB_2100179_Chapter5.pdf Download (242kB)
	Text S_MKB_2100179_Appendix.pdf Restricted to Staf Perpustakaan Download (5MB)

Item Type:	Thesis (S1)
Additional Information:	https://scholar.google.com/citations?hl=id&user=ysrR5ewAAAAJ ID SINTA Dosen Pembimbing: Mahmudah Salwa Gianti: 6779018 Dewi Indriati Hadi Putri: 6720737
Uncontrolled Keywords:	SNMP, PRTG, Monitoring Jaringan, Prediksi Downtime, Wilayah 3T SNMP, PRTG, Network Monitoring, Downtime Prediction, 3T Regions.
Subjects:	T Technology > T Technology (General)
Divisions:	UPI Kampus Purwakarta > S1 Mekatronika dan Kecerdasan Buatan
Depositing User:	Geralda Livia Nugraha
Date Deposited:	08 Sep 2025 04:17
Last Modified:	08 Sep 2025 04:17
URI:	http://repository.upi.edu/id/eprint/137999

Actions (login required)

View Item