Recent Advances in Environmental Weather Monitoring and Prediction System Using IoT and Multi-Model Progressive Dense Self-Attention: A Systematic Review

Abstract

Environmental weather monitoring and prediction systems are essential for applications such as agriculture, disaster management, and smart cities. Traditional numerical weather prediction models, while reliable, often struggle to capture fine-grained spatial and temporal dependencies and require high computational resources. The integration of Internet of Things (IoT) and Artificial Intelligence (AI) has enabled real-time data collection and advanced predictive analytics. This review focuses on IoT-based weather monitoring systems combined with deep learning and self-attention architectures, particularly the Multi-Model Progressive Dense Self-Attention Network. IoT sensors continuously gather environmental data such as temperature, humidity, rainfall, and wind speed, generating large-scale time-series datasets. While models like CNNs and LSTMs are widely used, they have limitations in capturing long-range dependencies. Transformer-based models and attention mechanisms address these issues by effectively modelling global spatial and temporal patterns. Hybrid architectures integrating CNNs, transformers, and graph-based methods have shown improved accuracy and robustness. Despite these advancements, challenges such as computational complexity, data heterogeneity, and scalability persist, highlighting the need for efficient and scalable prediction systems.