Artificial intelligence driven cognitive optimization and predictive analysis using blockchain privacy-based machine learning model

Abstract

According to the cognitive radio paradigm, spectrum sensing, decision-making, sharing, and mobility phases can be integrated to enable both authorised and unauthorised users to coexist in the radio-electric spectrum to the fullest extent possible. Using a blockchain privacy model and ML approaches in a B5G network, this study suggests a new approach to cognitive network optimisation and predictive analysis which stands for blockchain privacy-based transfer encoder neural networks. One further step is to optimise the network using binary swarm optimisation (BSMO). Bandwidth efficiency, throughput, forecast accuracy, and quality of service are the experimentally measured variables for a given number of channels and users. An innovative QoS-based optimisation phase and two separate decision-making procedures are proposed for usage in a proactive approach. One use of artificial neural networks (ANNs) is the prediction of future traffic loads for different radio access technologies (RATs), which employ different bands of the electromagnetic spectrum.