Citation

Ali, Farman and Ramzani, Ubaid Ur Rehman and Jayakanthan, M.L. and Roslee, Mardeni and Fadli, Mohd Zul and Ullah, Yasir and Jizat, Noorlindawaty Md. and Almas, Anum (2026) Achieving 45 % water savings and 12 % forecast accuracy gain in 5G-enabled smart irrigation using IoT-based open-source technologies. Sustainable Computing: Informatics and Systems, 50. p. 101306. ISSN 2210-5379

Text 1-s2.0-S2210537926000168-main.pdf - Published Version Restricted to Repository staff only Download (11MB)

Abstract

Efficient irrigation is essential for sustainable agricultural productivity and food security, especially amid increasing water scarcity and climate variability. Recent models have introduced Internet of Things (IoT)-based smart irrigation systems using fuzzy logic, Auto Regressive Integrated Moving Average (ARIMA) forecasting, machine learning (ML), and embedded control technologies. However, these models rely on complex algorithms that demand high computational resources, lack multi-sensor integration for comprehensive field monitoring, and are limited by network scalability, and power consumption. This work presents a 5G-enabled smart irrigation framework that achieves up to 45 % water savings and a 12 % gain in soil moisture forecast accuracy using a multi-sensor IoT-based open-source architecture. The system integrates soil moisture, temperature, humidity, rainfall, ultrasonic water level, and smoke sensors to enable dynamic environmental monitoring across agricultural fields. A threshold-based control logic, supported by evapotranspiration modelling and ARIMA forecasting, drives the automated irrigation mechanism. Wireless communication is established through the NodeMCU ESP8266 Wi-Fi module, with scalability toward 5 G networks for low-latency data transmission to a mobile application. The system ensures decision-making autonomy through calibrated sensor feedback and predictive environmental modelling. Across 12/24/48-hour horizons, the framework attains RMSE = 0.82/0.94/1.12 and NRMSE = 4.1 %/4.8 %/5.3 %, outperforming LoRaWAN, MQTT, and WDO/AWDO-LSSVM baselines. Edge/fog execution yields sub-second actuation latency, and a confidence-weighted fusion with model-based failover sustains decisions under sensor drift or dropouts. The design remains low-cost and solar-powered, supporting deployment in resource-constrained agricultural scenarios.

Item Type:	Article
Uncontrolled Keywords:	Smart irrigation, Multi-sensor fusion, Evapotranspiration based, control, ARIMA soil-moisture forecasting, Edge/fog computing, 5 G IoT, Precision agriculture, Water savings, Forecast accuracy
Subjects:	T Technology > TD Environmental technology. Sanitary engineering > TD201-500 Water supply for domestic and industrial purposes
Divisions:	Faculty of Artificial Intelligence & Engineering (FAIE)
Depositing User:	Ms Suzilawati Abu Samah
Date Deposited:	27 Feb 2026 07:57
Last Modified:	27 Feb 2026 07:57
URII:	http://shdl.mmu.edu.my/id/eprint/15363

Downloads

Edit (login required)