Deep learning can achieve higher accuracy than traditional machine learning algorithms in a variety of machine learning tasks. Recently, privacy-preserving deep learning has drawn tremendous attention from information security community, in which neither training data nor the training model is expected to be exposed. Federated learning is a popular learning mechanism, where multiple parties upload local gradients to a server and the server updates model parameters with the collected gradients. However, there are many security problems neglected in federated learning, for example, the participants may behave incorrectly in gradient collecting or parameter updating, and the server may be malicious as well. In this paper, we present a distributed, secure, and fair deep learning framework named \textit{DeepChain} to solve these problems. DeepChain provides a value-driven incentive mechanism based on Blockchain to force the participants to behave correctly. Meanwhile, DeepChain guarantees data privacy for each participant and provides auditability for the whole training process. We implement a DeepChain prototype and conduct experiments on a real dataset for different settings, and the results show that our DeepChain is promising.
19 November 2018 in IEEE Transactions on Parallel and Distributed Systems
Crowdsourcing systems which utilize the human intelligence to solve complex tasks have gained considerable interest and adoption in recent years. However, the majority of existing crowdsourcing systems rely on central servers, which are subject to the weaknesses of traditional trust-based model, such as single point of failure. They are also vulnerable to distributed denial of service (DDoS) and Sybil attacks due to malicious users involvement. In addition, high service fees from the crowdsourcing platform may hinder the development of crowdsourcing. How to address these potential issues has both research and substantial value. In this paper, we conceptualize a blockchain-based decentralized framework for crowdsourcing named CrowdBC, in which a requester's task can be solved by a crowd of workers without relying on any third trusted institution, users' privacy can be guaranteed and only low transaction fees are required. In particular, we introduce the architecture of our proposed framework, based on which we give a concrete scheme. We further implement a software prototype on Ethereum public test network with real-world dataset. Experiment results show the feasibility, usability and scalability of our proposed crowdsourcing system.
31 March 2020 in IEEE Transactions on Vehicular Technology
Load frequency control (LFC) is widely employed to keep smart grids stable and secure. This paper proposes an adaptive resilient LFC scheme for sub-systems of smart grids under denial-of-service (DoS) attacks with energy constraint. Firstly, a resilient triggering communication scheme is introduced, where the triggering condition includes the uncertainty item induced by DoS attacks. Secondly, an adaptive resilient event-triggering LFC scheme is proposed to further reduce the communication burden and defeat the DoS attacks, where the event-triggering parameter can be dynamically adjusted. Third, a stability criterion is derived for Proportional Integral-based LFC systems by employing Lyapunov theory. Finally, to validate the proposed control scheme, case studies are carried out based on three different LFC power generation systems including one-area, two-area and three-area models. The simulation results clearly demonstrate that the proposed adaptive resilient LFC scheme can reduce the communication burden while defeating DoS attacks by comparing with resilient event-triggering LFC scheme and event-triggering LFC scheme in term of defense effect and average triggering period.
9 November 2018 in IEEE Transactions on Vehicular Technology
Recently emerging SDN-based VANET (i.e., vehicular ad hoc network based on software-defined networking) enables VANET management to be programmable and flexible. It introduces SDN controllers to maintain network-wide resources and SDN applications to program configurations through arbitrarily accessing resources via the northbound interface (NBI). However, this brings with it security issues on the NBI, such as network-wide resource exposure and configuration manipulation. Most of the existing works employed permission systems to restrict resource access; these solutions are generally controller-dependent, which means controller codes need to be modified for giving access permissions to external applications. In this paper, we propose a scalable and dynamic access control scheme on the NBI for SDN-based VANET, named BENBI. In the proposed scheme, we dynamically and flexibly control network resources by employing Broadcast Encryption (BE), rather than altering source codes of the controller or updating permission lists with various degrees of granularity. Moreover, the resources are encrypted during transmission so that they are only available to authorized applications. Finally, we implement a prototype of BENBI. The experimental results demonstrate that the cost of allocating secret keys is independent of the number of SDN entities being appointed, which indicates the scalability of our scheme.