Relationship-Cluster Head Selection and Data Compression Enabled Cluster-Based Aggregation Model for Social Internet of Things

Abstract

Background:The Social Internet of Things (SIoT) combines IoT with social networking, enabling objects to interact based on social relationships and facilitating user-device interactions. However, SIoT networks generate massive data that must be transmitted and processed efficiently. We propose a clusterbased aggregation model for SIoT to address this, integrating relationshipcluster head selection with K-Means and data compression using Huffman coding. Objectives: The proposed model utilizes K-Means to select cluster heads based on object relationships. Objects are clustered using K-Means, and a Decision Tree considers object profiling and relationships to choose the cluster head. Selected cluster heads employ Huffman coding to compress data before transmission to the sink node. Methods: The data are aggregated at the cluster head and it is compressed before sending it to the destination device. Model is evaluated through simulation-based experiments using the SIoT-CCN simulator. Findings demonstrate that our proposed model outperforms existing approaches in terms of energy consumption, network lifetime, and data aggregation accuracy. Findings: Evaluation metrics include Silhouette Score, Number Distance between Train and Test K-Means, BIC Score, and Gradient of BIC Scores. Results for our proposed model include a Silhouette score of 0.459 for cluster numbers 2, 3, and 4, -1148.951 distance between train and test K-Means, -5227.080 BIC Score, and -96.445 Gradient of BIC score. For the Relation-Based clustering approach without data compression, the Silhouette score is 0.6266, -203.345 distance between Train and test K-Means, -7266.080 BIC Score, and -61.415 Gradient of BIC Scores. The Data Compression-enabled cluster-based aggregation model without relationshipbased clustering achieves a Silhouette score of 0.58, -467.890 distance between train and test K-Means, -8981.786 BIC Scores, and -94.244 Gradient of BIC Score. Novelty: Integrating relationship-cluster head selection with K-Means and data compression with Huffman coding to develop a cluster-based https://www.indjst.org/ 3605