Joint Protection of Energy Security and Information Privacy for Energy Harvesting: An Incentive Federated Learning Approach

Abstract

Energy harvesting (EH) is a promising and critical technology to mitigate the dilemma between the limited battery capacity and the increasing energy consumption in the Internet of everything. However, the current EH system suffers from energyinformation cross threats, facing the overlapping vulnerability of energy deprivation and private information leakage. Although some existing works touch on the security of energy and information in EH, they treat these two issues independently, without collaborative and intelligent protection cross the energy side and information side. To address the above challenge, this paper proposes a joint protection framework of energy security and information privacy for EH with an incentive federated learning approach. First, we design a federated learning-based malicious energy user detection method according to energy status and behaviors to provide energy security protection. Secondly, a differential privacy-empowered information preservation scheme is devised, where sensitive information is protected by the customized demand-based noise. Thirdly, a non-cooperative gameenabled incentive mechanism is established to encourage EH nodes to participate in the joint energy-information protection system. The proposed incentive mechanism derives the optimal energy-information security strategy for EH nodes and achieves a tradeoff between the protection of energy security and information privacy. Evaluation results have verified the effectiveness of our proposed joint protection mechanism.

Cyber security for federated learning environment using AI technique

“…Our testing was conducted out on a 4th‐generation Intel Core i5 processor with 12 GB of RAM organization Ubuntu 18.10. They utilized WEKA 3.8, an open‐sourced system which covers a variety of classifications and feature subset selecting approaches (Abawajy & Kelarev, 2019), like we did in (Ali et al, 2021; Hall et al, 2009; Pan et al, 2021; Yu et al, 2021). They was using an available to the public dataset with 6250 benign applications from of the Google Play Store and 5700 malicious applications from VirusShare, Drebin, and AndroZoo.…”

Software-Defined GPU-CPU Empowered Efficient Wireless Federated Learning With Embedding Communication Coding for Beyond 5G

“…Researchers have tried to optimize WFL from different aspects. Under the condition of limited wireless network resources [31] and client energy resources [32]- [33] participating in WFL local training, Zhou proposed a bandwidth allocation algorithm with low energy consumption [34], which enables clients to engage in learning more sustainably. Xu proposed to intelligently select clients participating in WFL local learning based on energy consumption from the longterm perspective of learning as a whole [35], not limited to learning rounds.…”

Intelligent Terrestrial and Non-Terrestrial Vehicular Networks with Green AI and Red AI Perspectives

“…Moreover, it is highly anticipated that the AI technologies including deep learning and machine learning will be driven by a wide range of innovative applications and systems such as 5G, edge computing, cloud, Internet of Things (IoT), tactile internet, wireless networks, vehicles, blockchain, etc. [ 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 ]. Furthermore, several standpoints are studied regarding AI technology and its applicability.…”