With the advances of machine learning algorithms and the pervasiveness of network terminals, the online medical prediagnosis system, which can provide the diagnosis of healthcare provider anywhere anytime, has attracted considerable interest recently. However, the flourish of online medical prediagnosis system still faces many challenges including information security and privacy preservation. In this paper, we propose an e fficient and privacy-preserving online medical prediagnosis framework, called eDiag, by using nonlinear kernel support vector machine (SVM). With eDiag, the sensitive personal health information can be processed without privacy disclosure during online prediagnosis service. Specifically, based on an improved expression for the nonlinear SVM, an efficient and privacy-preserving classification scheme is introduced with lightweight multiparty random masking and polynomial aggregation techniques. The encrypted user query is directly operated at the service provider without decryption, and the diagnosis result can only be decrypted by user. Through extensive analysis, we show that eDiag can ensure that users' health information and healthcare provider's prediction model are kept confidential, and has significantly less computation and communication overhead than existing schemes. In addition, performance evaluations via implementing eDiag on smartphone and computer demonstrate eDiag's effectiveness in term of real online environment.

Over the past years, the increasingly severe data island problem has spawned an emerging distributed deep learning framework—federated learning, in which the global model can be constructed over multiple participants without directly sharing their raw data. Despite its promising prospect, there are still many security challenges in federated learning, such as privacy preservation and integrity verification. Furthermore, federated learning is usually performed with the assistance of a center, which is prone to cause trust worries and communicational bottlenecks. To tackle these challenges, in this paper, we propose a privacy-preserving and verifiable decentralized federated learning framework, named PVD-FL, which can achieve secure deep learning model training under a decentralized architecture. Specifically, we first design an efficient and verifiable cipher-based matrix multiplication (EVCM) algorithm to execute the most basic calculation in deep learning. Then, by employing EVCM, we design a suite of decentralized algorithms to construct the PVD-FL framework, which ensures the confidentiality of both global model and local update and the verification of every training step. Detailed security analysis shows that PVD-FL can well protect privacy against various inference attacks and guarantee training integrity. In addition, the extensive experiments on real-world datasets also demonstrate that PVD-FL can achieve lossless accuracy and practical performance.

Similarity query over time series data plays a significant role in various applications, such as signal processing, speech recognition, and disease diagnosis. Meanwhile, driven by the reliable and flexible cloud services, encrypted time series data are often outsourced to the cloud, and as a result, the similarity query over encrypted time series data has recently attracted considerable attention. Nevertheless, existing solutions still have issues in supporting similarity queries over time series data with different lengths, query accuracy and query efficiency. To address these issues, in this article, we propose a new efficient and privacy-preserving similarity range query scheme, where the time warp edit distance (TWED) is used as the similarity metric. Specifically, we first organize time series data into a $k$ d-tree by leveraging TWED’s triangle inequality, and design an efficient similarity range query algorithm for the $k$ d-tree. Second, based on a symmetric homomorphic encryption technique, we carefully devise a suite of privacy-preserving protocols to provide a security guarantee for $k$ d-tree based similarity range queries. After that, by using the similarity range query algorithm and these protocols, we propose our privacy-preserving similarity range query scheme, in which we elaborate on two strategies to make our scheme resist against the cloud inference attack. Finally, we analyze the security of our scheme and conduct extensive experiments to evaluate its performance, and the results indicate that our proposed scheme is indeed privacy-preserving and efficient.

Federated learning (FL) has nourished a promising method for data silos, which enables multiple participants to construct a joint model collaboratively without centralizing data. The security and privacy considerations of FL are focused on ensuring the robustness of the global model and the privacy of participants’ information. However, the FL paradigm is under various security threats from the adversary aggregator and participants. Therefore, it is necessary to comprehensively identify and classify potential threats to provide a theoretical basis for FL with security guarantees. In this paper, a unique classification of attacks, which reviews state-of-the-art research on security and privacy issues for FL, is constructed from the perspective of malicious threats based on different computing parties. Specifically, we categorize attacks with respect to performed by aggregator and participant, highlighting the Deep Gradients Leakage attacks and Generative Adversarial Networks attacks. Following an overview of attack methods, we discuss the primary mitigation techniques against security risks and privacy breaches, especially the application of blockchain and Trusted Execution Environments. Finally, several promising directions for future research are discussed.

Online medical primary diagnosis system, which can provide convenient medical decision support through applying mobile communication and data analysis technology, has been considered as a promising approach to improve the quality of healthcare service. However, it still faces many severe challenges on the privacy of users’ health information and the accuracy of diagnosis result, which deter the wide adoption of online medical primary diagnosis system. In this paper, we propose an efficient and privacy-preserving online medical primary diagnosis framework (CINEMA). Within CINEMA framework, users can access online medical primary diagnosing service accurately without divulging their medical data. Specifically, based on fast secure permutation and comparison technique, the encrypted user’s query is directly operated at the service provider without decryption, and the diagnosis result can only be decrypted by the user, meanwhile, the diagnosis model in service provider can also be protected. Through extensive analysis, we show that CINEMA can ensure that user’s health information and healthcare service provider’s diagnosis model are kept confidential, and has significantly reduce computation and communication overhead. In addition, performance evaluations via implementing CINEMA demonstrate its effectiveness in term of the real environment.

With the prosperity of mobile internet and the pervasiveness of location-aware mobile terminals, online ride-hailing, a high-level location-based service (LBS) which relies on dynamic spatial query, has made our life more convenient. However, the flourish of ride-hailing service still faces many severe challenges since users' location privacy and service provider's data security. In this paper, we present an efficient and privacy-preserving dynamic spatial query scheme (TRACE) for ride-hailing service. With TRACE, users (i.e., consumers and vehicles) can access ride-hailing service without divulging their sensitive location information, meanwhile, the ride-hailing server can achieve the necessary commercial operating information while keeping its sensitive data (i.e., the space division information) confidential. Specifically, with two proposed efficient and secure spatial query algorithms, named FSSQ and ESVQ, all location-related data are encrypted by its owner before being sent out, and are calculated without decryption during the spatial query process. Therefore, consumers, vehicles, and service provider cannot obtain each other's sensitive information. Detailed security analysis shows that TRACE can resist various known security threats. Furthermore, TRACE is implemented in the real environment, and extensive simulation results over smart phones demonstrate that the scheme is highly efficient and can be implemented effectively.