Predicting the Performance of Internet of Things Communication Systems Using Finite Element Interpolated Neural Network in Smart Cities

The Internet of Things (IoT) is an essential part of Information and Communications Technology (ICT) for sustainable smart cities because of its capacity to assist sustainability across multiple disciplines. To attain the required quality of IoT communication systems and to enable sustainable progress in smart cities regarding IoT communication systems, it is necessary to avoid fault through constant and dynamic application of network behavior. In this research work, predicting the performance of IoT communication systems using Finite Element Interpolated Neural Network in smart cities (IoT-CS-FEINN-SC) is proposed. Here, the input data is gathered from IoT devices that include various kinds of sensors like visibility, humidity, temperature, pressure, and wind speed. Signed Cumulative Distribution Transform (SCDT) is employed to extract Received Signal Strength (RSS) features as minimum, maximum, and mean. Afterwards, the extracted features are fed to FEINN for predicting the IoT communication system performance in smart cities. The Secretary Bird Optimization Algorithm (SBOA) is proposed to enhance the weight parameter of FEINN method that predicts the performance of IoT communication systems precisely. The IoT-CS-FEINN-SC technique achieves 20.36%, 28.42%, and 15.27% better accuracy analyzed with existing techniques: Cloud-assisted IoT intelligent transportation scheme and traffic control scheme in smart city (IoT-TCS-SC), Optimized RNN-dependent performance prediction of IoT and WSN-oriented smart city application utilizing improved honey badger algorithm (RNN-IoT-WSN), and Smart cities: a role of IoT and ML in realizing data-centric smart environs (IoT-ANN-DSE), respectively.