• B. Gnaneshwari

The Internet of Things (IoT) ecosystem faces significant privacy and security challenges due to its heterogeneous devices, dynamic behavior, and evolving hacking techniques. Traditional approaches struggle to provide a unified solution. To address these issues, this study proposes an integrated privacy and security mechanism using a novel encoding technique known as Replacement Encoding (RE). Designed to protect sensitive information during AI model training, RE camouflages data while preserving model integrity and utility. Additionally, it offers automated data preprocessing to enhance AI performance. Message packet features were extracted from the CICIoT2023 dataset (PCAP files) using Wireshark for analysis. These features were processed with the proposed RE technique and integrated into AI classifiers for attack detection. The RE-based models achieved accuracies of 88.94% using Random Forest (RF) and 86.61% using Deep Neural Networks (DNN), employing 100 features. In comparison, models utilizing genetic algorithm-based correlation features with up to 15 features reached accuracies of 90.16% (RF) and 94.81% (DNN), indicating the potential of optimized feature selection. The study demonstrates the RE mechanism’s effectiveness not only in maintaining privacy and enhancing preprocessing but also in securing sensitive data. Its applicability extends to various domains, including Generative Pre-Training Transformer (GPT) systems and other AI applications. Overall, this unified framework contributes significantly to the advancement of secure, intelligent IoT systems.

AI, encoding, IoT, privacy, optimization, security

" Enhancing Privacy, Security, and Performance in IoT Ecosystems Using AI-Powered Encoding Mechanisms", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 5, page no.508-522, May-2025, Available :http://www.jetir.org/papers/JETIRGV06073.pdf

" Enhancing Privacy, Security, and Performance in IoT Ecosystems Using AI-Powered Encoding Mechanisms", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 5, page no. pp508-522, May-2025, Available at : http://www.jetir.org/papers/JETIRGV06073.pdf