A Multi-Layered Security Approach: Protecting MQTT Protocols, Enhancing Sensor Data via DWT, and Safeguarding Cloud Data Integrity
Main Article Content
Abstract
The paper examines novel approaches to improve the security, effectiveness, and fault diagnosis of Internet of Things systems, with an emphasis on MQTT protocol security and discrete wavelet transform (DWT) signal processing. Among the main concerns addressed are data integrity problems in cloud storage, security flaws in MQTT-based communications, and noise reduction in various IoT contexts. Lebesgue Hash Message Double Authentication fused Deep Learning (LHMDA-DL) is a hybrid deep learning technique that is suggested to protect MQTT message transport from malicious variations and Denial-of-Service (DoS) attacks. It uses Hellinger Distributed Stochastic Neighbor Embedding for data preprocessing, Gaussian Mixture Models with Lebesgue measures for feature extraction, and Hash Message Double Authentication Rules for secure message transmission over Long Short-Term Memory (LSTM) networks. Through simulation, the efficacy of the framework is confirmed, demonstrating notable gains over current methods in processing time, traffic overhead reduction, message transmission accuracy, and error rate. In order to reduce risks and maintain user data security, sophisticated encryption techniques and hierarchical key management are highlighted as ways for guaranteeing data integrity in cloud storage systems. This thorough investigation shows how deep learning models, IoT security standards, and sophisticated wavelet signal analysis work together to create a strong foundation for safe and effective cloud and IoT operations.