IoT-Enabled Traffic Management Systems using CNN-TransLSTM for Next-Generation Smart Cities

The rising quantity of vehicles has intensified traffic congestion, pollution, and road accidents. This paper introduces an IoT-enabled traffic management system utilizing CNN-TransLSTM, a hybrid model that combines convolutional neural networks, LSTM, and transformers for effective traffic prediction. Min-max normalization eliminates outliers and maintains data integrity, whereas feature extraction identifies essential variables such as weather, traffic density, and direction. The model attained a prediction accuracy of 91.25%, exceeding that of individual CNN, LSTM, and Transformer models. This method emphasizes the capacity of smart cities to utilize deep learning and IoT for superior traffic control, resulting in increased efficiency and sustainability.