Advanced IoT Intrusion Detection for Intelligent Homes using Optimized Cross-Contextual Transformers with a Dynamic City Game Framework

Ensuring data security against unauthorized access in smart homes is crucial, but implementing deep learning methods can be challenging due to potential issues such as accuracy loss, time complexity, and increased error rates. This study introduces a novel approach to address these challenges: the Securing Intelligent Home with a Transformer-Based IoT Intrusion Detection System (IDS) using an Optimized Cross-Contextual Point Running City Game Transformer Network (O2CPRCG-TransNet). This innovative method leverages advanced DL approaches to improve the security and reliability of Intelligent home environments. The study utilizes the IoT_bot dataset, which is pre-processed using the Grid-Constrained Data Purification Method (GCDP) to ensure high-quality input. The purified data is then optimized through the Giant Trevally Optimizer Algorithm (GTOA), which selects the most significant features to improve classification performance. Classification is performed using the O2CPRCG-TransNet architecture, which is specifically designed to address the complexities of Intelligent home security. To ensure secure data transmission, the system incorporates Somewhat Homomorphic Fuzzy-based Elliptic Curve Cryptography (SWH-FECC). Experimental results, obtained using the Python platform, demonstrate that the proposed approach achieves superior performance, with an accuracy of 98% and a recall of 0.993, compared to existing methods. These findings underscore the effectiveness of the proposed method and its potential for broader application in enhancing Intelligent home security.