IOT architecture for smart cities

Abstract

The proliferation of users in IoT applications has led to a drastic newlineamplification in data production. Among these applications, smart city newlineapplications stand out as complex systems, incorporating various sub newlineapplications such as smart healthcare, smart grid, smart parking, and newlineenvironmental monitoring. Various architectures have been designed for newlinedifferent smart city applications to overcome challenges such as newlineinteroperability, communication costs, latency, and network usage. newlineIn the context of integrating various devices in IoT and supporting newlineinteroperability, a three-tier middleware architecture is proposed. This newlinearchitecture is specifically applied in smart healthcare, particularly in real newlinetime pregnancy monitoring applications aimed at mitigating mortality rates. newlineAn analytical platform integrates IoT sensors such as EMG and ECG to newlineanalysing vital parameters like contractions, fetal heart rate, and fetal position. newlineSignal processing is executed at the analytic-as-a-service platform, and newlineabnormalities in fetal heart rate, fetal position, and distinguishing actual labor newlinefrom prodromal (false) labor are notified to medical advisors or family newlinemembers through the proposed solution. newlineAnother significant challenge in smart city applications is the increase newlinein data generation at the edge. As the workload of cloud escalates, resulting in newlinenetwork overhead, latency, and increased communication costs, a Hadoop newlinebased architecture is proposed and deployed at the network edge. Hadoop, newlineknown for its efficiently handling big data processing in the cloud, is utilized newlineat the edge. This proposed architecture enhances the data processing speed at newlinethe edge and drastically reduces data transfer to the cloud, thereby reducing newlinecommunication costs and cloud energy consumption by 91.4%. newline