University of Sussex
Browse
electronics-11-03339.pdf (1.39 MB)

Blockchain-empowered AI for 6G-enabled internet of vehicles

Download (1.39 MB)
journal contribution
posted on 2023-06-10, 05:08 authored by Ferheen AyazFerheen Ayaz, Zhengguo ShengZhengguo Sheng, Daxin Tian, Maziar NekoveeMaziar Nekovee, Nagham Saeed
The 6G communication technologies are expected to provide fast data rates and incessant connectivity to heterogeneous networks, such as the Internet of Vehicles (IoV). However, the resulting unprecedented surge in data traffic, massive increase in the number of nodes with high mobility, and low-latency requirements give rise to serious security, privacy, and trust challenges. The blockchain could potentially ensure trust and security in IoV due to its features, including consensus for credibility and immutability for tamper proofing. In parallel, federated learning (FL) is a privacy-preserving artificial-intelligence paradigm that does not require to share data for model training in machine learning. It can reduce data traffic and resolve privacy challenges of intelligent IoV networks. The blockchain can also complement FL by ensuring the decentralization and securing distribution of incentives. This article reviews the trends and challenges of the blockchain and FL in 6G IoV networks. Then, the impact of their combination, challenges in implementation, and future research directions are highlighted. We also evaluate our proposal of blockchain-based FL to protect IoV security and privacy that utilizes smart contract and secure transactions of incentives via the blockchain to protect FL. Compared with other solutions, the failure rate of the proposed solution was at least 5% lower with 30% malicious nodes in the network.

History

Publication status

  • Published

File Version

  • Published version

Journal

Electronics

ISSN

2079-9292

Publisher

MDPI AG

Issue

20

Volume

11

Page range

3339 1-12

Department affiliated with

  • Engineering and Design Publications

Full text available

  • Yes

Peer reviewed?

  • Yes

Legacy Posted Date

2022-10-18

First Open Access (FOA) Date

2022-10-18

First Compliant Deposit (FCD) Date

2022-10-17

Usage metrics

    University of Sussex (Publications)

    Categories

    No categories selected

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC