Trust-Aware Secure Communication Framework for IoT Networks: A Hybrid Approach Integrating Instantaneous Trust and Dynamic Key Management

The increasing prevalence of cyber threats such as Man-in-the-Middle (MITM) attacks, data breaches, and unauthorized access in Internet of Things (IoT) networks necessitates the development of robust, adaptive security frameworks. Existing approaches to IoT security, including trust-based authentication, proxy re-encryption, and lightweight cryptographic mechanisms, have limitations in dynamic threat environments. This paper proposes the Trust-Aware Secure Communication Framework (TASCF-IoT), a novel hybrid approach that integrates instantaneous trust calculation using communication channel characteristics with a resource-aware dynamic key-sharing mechanism to ensure secure and resilient IoT communication. The framework introduces a Trust-Enhanced Proxy Re-Encryption (TE-PRE) scheme that dynamically adjusts encryption policies based on real-time trust assessment, mitigating security risks in IoT-fogcloud ecosystems. Additionally, a novel anomaly detection model leveraging Long Short-Term Memory (LSTM)-based delay prediction and Hidden Markov Models (HMMs) is incorporated to identify suspicious behaviors and mitigate active MITM attacks with high accuracy. This method is validated using both a simulated testbed and real-world IoT environments. The experimental results demonstrate that TASCF-IoT achieves an attack detection accuracy (ADA) of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 9. 2 \%}$</tex>, outperforming conventional intrusion detection systems and trust evaluation models. Furthermore, the proposed framework significantly reduces computational overhead while maintaining robust security in constrained IoT devices. The findings highlight the effectiveness of instantaneous trust-aware key management and predictive security modeling, offering a scalable and adaptive solution for next-generation IoT networks.