Efficient Artificial Intelligence-based Constrained Application Protocol (CoAP) Protocol for IoT-Enabled Machine Learning Decision and Security System
DOI:
https://doi.org/10.63075/3a0atg40Keywords:
Machine Learning, Deep Neural Network, CNN, Prediction models, Internet Of Things, Threat Detection, Internet Of Things Networks, Wi-Fi SecurityAbstract
The acquisition of data happens through devices with software-based internet capabilities. The Constrained Application Protocol (CoAP) provides application layer communication services for resource-constrained devices in current use. The current lightweight data protection approaches continue to experience attacks and require high-speed network links. The novel Artificial Intelligence-based (AI) hybrid lightweight data security method with CoAP enhancements functions as FPL-GLCoAP for providing security in IoT systems. The method consists of three essential functions, which include registration and authentication alongside lightweight data security. A new device obtains its credential registration as its first step when joining a network. Feister Substitution Permutation Block Cipher establishes the authentication process as part of its AI-based authentication mechanism. We have implemented the Feister Substitution Permutation Block Cipher as a lightweight AI data security mechanism that uses small-sized unique keys at each round to reduce data transmission to the cloud and preserve real-time analysis while saving bandwidth. The Light Weight Data Security (LWDS) receives evaluation through cryptanalysis that performs diffusion property tests (Galois Field multiplication, one-to-one matrix linear permutation within the diffusion layer). The main objective of incorporating Galois Field multiplication, one-to-one matrix linear permutation within the diffusion layer, is to safeguard the statistical relationships between plain text and ciphertext. Analysis confirms that proposed CoAP achieves higher accuracy together with data confidentiality, while reducing latency through minimal bandwidth utilization across the DS2OS traffic traces dataset. The proposed hybrid lightweight data security detection method provides enhanced performance results. Proposed CoAP exhibits better precision and accuracy levels of 13% and 15%, respectively as compared to CNN and ANN and SCOAP protocols..
