RPFSA_LeNet‐Enabled Routing and LeNet‐Based Big Data Classification in IoT Environment

ABSTRACT The Internet of things (IoT) generates an enormous amount of data from systems, sensors, and devices interlinked through the internet. Big data systems are important to handle and process these greater volumes of data effectually. Big data often include immense and different databases from several domains. The classification aids in arranging the data into consequential categories that make it simpler to retrieve significant information rapidly, which is also important for risk analysis in ensuring data security. Hence, this paper implements the Red Panda Flamingo Search Algorithm–enabled LeNet (RPFSA_LeNet) for the classification of big data in IoT. Here, IoT simulation is performed in the first stage. Then, energy prediction is performed by the Long‐Short Term Memory (LSTM). Following this, data are transmitted to the base station (BS) through the routing process, which is performed by RPFSA wherein fitness parameters are employed. Moreover, RPFSA is merged with Red Panda Optimization (RPO) and Flamingo Search Algorithm (FSA). Then, the Deep Embedded Clustering (DEC) performs the data partitioning. Big data classification is done after transmitting the data to BS. The MapReduce framework (MRF) is considered for classifying the big data. In the mapper phase, data are normalized using quantile normalization. The Wave‐Hedge distance is exploited for feature selection, which is then given to the reducer phase. Finally, LeNet does the big data classification, which is trained by the RPFSA. The measures like distance, delay, and energy exploited for routing gained 37.713 m, 0.565 s, and 0.037 J, and precision, recall, and F measure used for classification reached 91.931%, 91.476%, and 91.703%.