Authors: Samarth Upadhyay
Abstract: The rapid growth of Internet of Things (IoT) deployments has intensified the demand for architectures capable of processing high-velocity data streams and enabling real-time decision making. Traditional centralized cloud models are often inadequate for meeting the strict latency, scalability, and reliability requirements of modern IoT applications such as smart cities, industrial automation, healthcare monitoring, and autonomous systems. Cloud-native architectures, built on microservices, containerization, orchestration, and serverless computing, have emerged as a foundational paradigm for addressing these challenges. This review paper presents a comprehensive analysis of cloud-native architectures that support real-time IoT data processing and decision making. It systematically examines IoT system fundamentals, cloud-native design principles, streaming data pipelines, edge–cloud collaboration models, and decision-making mechanisms ranging from rule-based engines to machine learning–driven intelligence and digital twins. The paper further reviews data management strategies, performance evaluation metrics, and critical security and privacy considerations in distributed IoT environments. By synthesizing existing architectural approaches and comparative studies, this review identifies key design trade-offs, limitations, and research gaps, including challenges related to latency management, interoperability, system complexity, and trust. Finally, the paper outlines future research directions such as AI-driven self-adaptive architectures, edge intelligence, federated learning, and integration with next-generation networks. The findings provide valuable insights for researchers and practitioners seeking to design scalable, resilient, and intelligent cloud-native IoT systems capable of supporting real-time decision making.
DOI: http://doi.org/10.5281/zenodo.18170303
