Utilizing big data analytics for the sustainable development of smart cities on the IoT platform

In a data-driven environment, many companies generate and evaluate vast volumes of data to derive benefits. A smart city (SC) exemplifies using big data to enhance services for residents and tourists. Nevertheless, several nations encounter specific challenges in analysing big data incorporation for sustainability in the smart development of cities. This study paper aims to determine and examine the major obstacles to sustainable smart city (SSC) development. This study introduced big data analytics for smart cities' sustainable development on the IoT platform (BDA-SSC-IoT). In a SC context, an array of sensors, including smart parking sensors, intelligent home sensors, vehicular networking, water and temperature sensors, and object tracking devices, has been deployed. Feature Selection (FS) has shown to be an effective data purification approach for reducing redundant data and improving system efficiency in Deep Learning (DL). This research proposes an effective Deep Q-network (DQN) based FS approach for data (from multiple sources) cleaning, influenced by Reinforcement Learning (RL), which acquires information via experience. Extensive assessments have been performed on various types of IoT-BDA within an SSC framework to validate the efficacy and worth of the proposed approach.