The ISC International Journal of Information Security

Abstract In the realm of machine learning, Generative Adversarial Networks (GANs) have revolutionized the generation of synthetic data, closely mirroring the distribution of real datasets. This paper delves into the privacy concerns associated with GANs, particularly focusing on Membership Inference Attacks (MIAs), which aim to determine if a specific record was used in training a model. Such attacks pose significant privacy risks, especially when sensitive data is involved. To combat this, we propose a novel detector model designed to identify and thwart MIAs within GANs. Our model, which operates as an additional layer of protection for Machine Learning as a Service (MLaaS) providers, leverages outputs from both the discriminator and generator to ascertain the membership status of data samples. We introduce two variants of the detector model—supervised and unsupervised—based on the availability of information from the discriminator. The supervised detector employs labeled data for training, while the unsupervised detector uses anomaly detection techniques. Our experimental evaluation spans various GAN architectures and datasets, ensuring the robustness and generalizability of our approach. The paper also analyzes the impact of dataset size on the detector's effectiveness. By integrating our detector, MLaaS providers can enhance privacy safeguards, striking a balance between model utility and data protection.

Abstract In recent years, blockchain technology has been used in many fields, including IoT and Smartphones. Since most of these devices are battery constrained and have low processing capabilities, conventional blockchains are not suitable for these types of systems. In this field, critical challenges that need to be addressed are providing security for transactions and power consumption. An available solution to meet the mentioned challenges is TrustChain. Unlike conventional blockchains, TrustChain does not have a single global chain. Instead, each node is responsible for building and maintaining its local chain.
With all the benefits, TrustChain is vulnerable to the whitewashing attack and suffers from client vulnerability issues. Moreover, once a fatal error occurs, the recovery time of each TrustChain node is considerably high. In this paper, we
propose a solution to address the attacks mentioned above by implementing an authentication system with MongoDB on top of TrustChain. Moreover, we connected TrustChain to the distributed cloud storage to significantly reduce the recovery time of nodes in fatal errors (up to 80%). Finally, we evaluate improved TrustChain with the PoW-based smartphone-oriented blockchains from two aspects of security and power consumption, proving that improved TrustChain does not significantly affect the lifetime of the smartphone battery. Its power consumption is less than mentioned blockchains and is more secure
than these systems against main attacks.