Bayesian Neural Networks for 6G CONASENSE Services

• Published in: 2022 25th International Symposium on Wireless Personal Multimedia Communications (WPMC) pp. 291 - 296
• Main Authors: Rufino Henrique, Paulo Sergio, Prasad, Ramjee
• Format: Conference Proceeding
• Language: English
• Published: IEEE 30-10-2022
• Subjects: 6G mobile communication; Bayesian Networks; Bnets; CNSS; Computational modeling; CONASENSE; Machine learning; Neural networks; QoS; Quality of service; Quantum computing; Quantum Machine Learning; Quantum mechanics; Society 5.0
• ISSN: 1882-5621
• DOI: 10.1109/WPMC55625.2022.10014760

CONASENSE is the research area that aims to foster the integration of communications, navigation, sense, and services. The existing independence of each CONASENSE area is the main challenge for its full integration. Therefore a chain of rules to model CONASENSE is needed to offer precision in controlling these areas with the aid of Artificial Intelligence (AI). The objectives are to create a causality model expressing the CONASENSE Networks (CNSS). Furthermore, CNSS must be mathematically and graphically fit to enable seamless functionality amongst its nodes. Simulating the plethora of probabilistic patterns that may occur in the complexity of uncertainties generated in these four areas is crucial. Thus, applying probabilistic distribution that tracks a dynamic variety of events conditional to each CNSS node in time progression is a tremendous advantage for efficiently designing and controlling its Quality of Service (QoS). In this perspective, belief networks, also known as the Bayesian Networks (Bnets), can fit very well to assist the proposed acyclic CNSS graphic model here suggested. It will also help comprehend how to accelerate such integration by 2030 and support the 6G roadmap. Then, the authors proposed a CONASENSE Bayesian Neural Network (CBNN) model to enable the realization of CNSS in the future 6G Cognitive Radio.