A decentralised blockchain-based air pollution spikes monitoring framework over intelligent IoT edge networks

Abstract

Air pollution spikes pose significant challenges to public health, environmental stability, and economic development, demanding robust and innovative solutions for effective monitoring and management. This research introduces a decentralized blockchain-based framework for air pollution spike monitoring, leveraging intelligent IoT edge networks to enhance data accuracy, reliability, and timeliness. By integrating blockchain technology with IoT sensors and edge computing, the proposed framework aims to overcome the limitations of traditional air quality monitoring systems, which often struggle with centralized data processing, security vulnerabilities, and delayed response times. The framework employs IoT devices with multi-pollutant sensors to capture real-time air pollution spike data. These edge devices preprocess and analyze the collected data locally, reducing the need for constant communication with centralized servers and thus enhancing energy efficiency and responsiveness. The edge network architecture ensures that critical information is processed and acted upon promptly, enabling immediate detection and mitigation of pollution spikes. Blockchain technology is integrated into the framework to ensure the collected data's integrity, transparency, and security. Utilizing a decentralized ledger, the system records all air quality data and transactions immutably, preventing data tampering and fostering trust among stakeholders, including regulatory authorities, industries, and the general public. Smart contracts are deployed on the Ethereum platform to automate the enforcement of air quality regulations, issuing fines to polluters who exceed predefined emission thresholds. This automation streamlines regulatory compliance and enhances accountability and enforcement efficiency. Machine learning algorithms are applied to the gathered data to predict future pollution trends and detect anomalies in real time. The research compares various time series models, including exponential smoothing, ARIMA, and Recurrent Neural Networks (RNNs), to determine the most effective approach for forecasting air pollution spikes. The exponential smoothing model demonstrated superior performance in terms of accuracy (with Mean Absolute Error (MAE) and Mean Square Error (MSE) of 3.66 and 84.86 respectively for PM2.5) and computational efficiency (due to its simple recursive calculations, making them faster to compute and requiring less memory), enabling reliable prediction and proactive management of pollution events.

xx | P a g e This research signifies a paradigm shift in air quality management by combining advanced technologies to create a comprehensive, decentralized monitoring system. The findings underscore the potential of blockchain and intelligent IoT edge networks to revolutionize the way air pollution is monitored and managed, ultimately contributing to improved public health outcomes and environmental sustainability. The results of this research demonstrated the successful design of a prototype framework capable of collecting air pollution spikes while conserving energy for more than 40% by activating only when peaks exceed a predefined threshold. The designed framework demonstrated also that by waking up using the threshold measurement, the possibility of losing some spikes which was 7.1% is no longer there as shown using Monte Carlo. The collected data were analyzed using blockchain technology to design smart contracts for air pollution spikes, ensuring efficient and cost-effective implementation. Additionally, the data were analyzed using machine learning models to predict future spikes. Among the three models evaluated (exponential smoothing, RNNs, and ARIMA), exponential smoothing outperformed the others, proving to be the most effective for this application for all pollutants