Blockchain-based Deep Learning Management to Enable Smart NextG Wireless Networks
Rui Ning, research assistant professor, cybersecurity, Old Dominion University
Hongyi Wu, Batten Chair Professor, director, School of Cybersecurity, electrical and computer engineering, Old Dominion University
Chunsheng Xin, professor, electrical and computer engineering, Old Dominion University
Hongzhi Guo, assistant professor, engineering, Norfolk State University
Blockchain-based Deep Learning Management to Enable Smart NextG Wireless Networks Project Description:
NextG networks are envisioned to offer full automation with connected intelligent devices seamlessly working together. To realize this vision, artificial intelligence (AI) and especially deep learning (DL) have emerged as critical technologies for networks that are too dynamic and complex to be analyzed by human operators. The convergence of deep learning and NextG will provide data-driven tools to support computational radio and network intelligence from the physical layer to network management.
Although deep learning has demonstrated its superior performance in many applications, it faces several fundamental challenges when it is applied to support the NextG wireless networks. In these NextG networks, deep learning models must be trained using multi-modality data from a massive amount of distributed devices in order to make comprehensive and collaborative predictions. The traditional centralized deep learning approach has limitations such as privacy concerns, data poisoning and data security. In this project researchers plan to address these technical challenges by developing a novel deep learning management framework based on blockchain and cryptographic primitives to enable secure and pervasive deep learning in the NextG wireless networks.