Security and privacy of machine learning assisted P2P networks
- PDF / 150,697 Bytes
- 3 Pages / 595.276 x 790.866 pts Page_size
- 29 Downloads / 207 Views
Security and privacy of machine learning assisted P2P networks Hongwei Li 1 & Rongxing Lu 2 & Mohamed M. E. A. Mahmoud 3
# Springer Science+Business Media, LLC, part of Springer Nature 2020
Advances in machine learning (ML) in recent years have enabled a dizzying array of applications in diverse areas of networks and communications. Specifically, the development of Peer-to-Peer (P2P) networks is promoted by either traditional or most advanced ML techniques in terms of efficiency, functionality as well as the scalability. Examples of such promotions are easy to find. As typical functions in P2P networks, the performance of neighbor selection, traffic classification and resource management could be significantly improved with the help of ML. Furthermore, with the rapid growing of network nodes and the network traffic, the diversity and the vast volume of the data also poses great challenges to the management and QoS of network. Therefore, ML cloud be one of the most promising novel technologies to resolve or mitigate such challenges. Although remarkable and promising benefits are brought through leveraging ML techniques to solve the problems in P2P networks, the security and privacy issues are becoming one of the most troublesome barriers. On one hand, personal data leakage, neighbor selection attacks, system penetration, Botnet attacks, and numerous traditional security and privacy threatens continuously hinder the development of the P2P network. Fortunately, ML is already proved to be one of the practical techniques to resist such attacks. On the other hand, the ML-assisted P2P network may introduce new security and Guest Editors: Hongwei Li, Rongxing Lu and Mohamed Mahmoud * Hongwei Li [email protected] Rongxing Lu [email protected] Mohamed M. E. A. Mahmoud [email protected] 1
University of Electronic Science and Technology of China, Chengdu, China
2
University of New Brunswick, Fredericton, NB, Canada
3
Tennessee Technological University, Cookeville, TN, USA
privacy issues. Numerous previously conducted researches have shown that the newly trained models may suffer from intensive black box and white box attacks. From the distribution of the training data set to the model parameters, numerous information could be the target of internal and external attackers. Therefore, such problems still exist in the MLassisted P2P networks. Moreover, some new and unexplored security and privacy problems also deserve further study. Hence, it is urgent to conduct researches to address security and privacy issues in ML-assisted P2P networks. This special issue has gained overwhelming attention and received 37 submissions from researchers and practitioners working on security and privacy of ML-assisted P2P networks. Each paper is selected into the regular rigorous review process and each submission has been reviewed by at least three reviewers. After 2–3 round reviews, eventually nine quality papers are recommended to be included into this special issue, which are summarized as below. The paper titled “CSNN: Password Guess
Data Loading...
