Core Concepts
This paper presents SWI-FEED, a comprehensive framework that integrates water distribution system simulation, wireless network simulation, and optimization algorithms to enable holistic assessment and optimization of smart water distribution systems.
Abstract
The paper introduces SWI-FEED, a framework designed to facilitate the widespread deployment of the Internet of Things (IoT) for enhanced monitoring and optimization of Water Distribution Systems (WDSs). The framework aims to investigate the utilization of massive IoT in monitoring and optimizing WDSs, with a particular focus on leakage detection, energy consumption, and wireless network performance assessment in real-world water networks.
The key components of the framework are:
Water Distribution System (WDS) Network Topology and Features:
The framework utilizes the WNTR/EPANET tools to analyze the WDS network topology and generate a dataset containing hydraulic features (e.g., demand, pressure, leakage) and wireless network features (e.g., data rate, energy consumption, node positions).
Wireless Network Analysis:
The framework integrates the NS-3 network simulator to analyze the performance of the LoRaWAN network used to collect data from the WDS, including parameters such as energy consumption, node battery lifetime, and network coverage.
Optimization Algorithms and Sustainable Applications:
The framework applies various optimization algorithms, including machine learning models, to the hydraulic and wireless network data to enable sustainable applications such as leakage detection, energy-efficient sensor and gateway deployment, and flow reconstruction.
The paper presents a use case demonstrating the framework's capabilities, focusing on the evaluation of strategies for LoRaWAN gateway deployment in alignment with the WDS hydraulic flow. The results show that the proposed Degree Centrality Deploy method consistently exhibits lower energy consumption compared to a Regular Grid Deploy approach, highlighting the potential benefits of using network centrality measures for gateway placement optimization in smart water distribution systems.
Stats
The daily network energy consumption for the Regular Grid Deploy and Degree Centrality Deploy methods with different numbers of gateways:
77 gateways:
Regular Grid Deploy: 189574 J
Degree Centrality Deploy: 147500 J
96 gateways:
Regular Grid Deploy: 161589 J
Degree Centrality Deploy: 131959 J
117 gateways:
Regular Grid Deploy: 116585 J
Degree Centrality Deploy: 98625 J
140 gateways:
Regular Grid Deploy: 96648 J
Degree Centrality Deploy: 82626 J
165 gateways:
Regular Grid Deploy: 66621 J
Degree Centrality Deploy: 60443 J
Quotes
"As the density of GWs increases, the ADR assigns lower SF values to the end devices. However, it is evident that the proposed method consistently exhibits lower energy consumption than Regular Grid Deploy in all configurations."