Hybrid butterfly optimization and back propagation neural network for enhanced smart city data classification

Smart cities aim to enhance resource efficiency and improve citizens’ quality of life through advanced technologies. One of the key challenges in this domain is accurately classifying large volumes of heterogeneous and imbalanced data generated by urban systems such as traffic, pollution, and utilities. This paper presents a novel hybrid classification framework that integrates the Butterfly Optimization Algorithm (BOA) with a Back Propagation Neural Network (BPNN), referred to as HBPNNBO, for enhanced smart city data classification. The framework begins with data preprocessing using the HADASYNBSID technique for balancing the dataset, followed by feature selection through a Hybrid Chicken Swarm Genetic Algorithm (HCSGA). To ensure secure and decentralized data handling, the framework incorporates blockchain technology coupled with a hybrid AES-CSO encryption method. This integration ensures end-to-end data integrity and privacy during classification tasks. The proposed HBPNNBO model is evaluated using benchmark smart city datasets, including intrusion and traffic data, and compared against conventional classifiers. The results demonstrate that the recommended strategy achieves superior performance, with a classification improvement of 94.76% exactness and minimal processing time of 23.62 ms. The findings confirm that the HBPNNBO framework, enhanced with blockchain-based security, is well-suited for real-time, secure smart city data analytics.