Search

Differential privacy (DP), as a rigorous mathematical definition quantifying privacy leakage, has become a well-accepted standard for privacy protection. Combined with powerful machine learning techniques, differentially private machine learning (DPML) is increasingly important. As the most classic DPML algorithm, DP-SGD incurs a significant loss of utility, which hinders DPML's deployment in practice. Many studies have recently proposed improved algorithms based on DP-SGD to mitigate utility loss. However, these studies are isolated and cannot comprehensively measure the performance of improvements proposed in algorithms. More importantly, there is a lack of comprehensive research to compare improvements in these DPML algorithms across utility, defensive capabilities, and generalizability. We fill this gap by performing a holistic measurement of improved DPML algorithms on utility and defense capability against membership inference attacks (MIAs) on image classification tasks. We first present a taxonomy of where improvements are located in the machine learning life cycle. Based on our taxonomy, we jointly perform an extensive measurement study of the improved DPML algorithms. We also cover state-of-the-art label differential privacy (Label DP) algorithms in the evaluation. According to our empirical results, DP can effectively defend against MIAs, and sensitivity-bounding techniques such as per-sample gradient clipping play an important role in defense. We also explore some improvements that can maintain model utility and defend against MIAs more effectively. Experiments show that Label DP algorithms achieve less utility loss but are fragile to MIAs. To support our evaluation, we implement a modular re-usable software, DPMLBench, which enables sensitive data owners to deploy DPML algorithms and serves as a benchmark tool for researchers and practitioners., Comment: To appear in the ACM Conference on Computer and Communications Security (CCS), November 2023, Tivoli Congress Center, Copenhagen, Denmark

The passive approach to quantum key distribution (QKD) consists of removing all active modulation from the users' devices, a highly desirable countermeasure to get rid of modulator side-channels. Nevertheless, active modulation has not been completely removed in QKD systems so far, due to both theoretical and practical limitations. In this work, we present a fully passive time-bin encoding QKD system and report on the successful implementation of a modulator-free QKD link. According to the latest theoretical analysis, our prototype is capable of delivering competitive secret key rates in the finite key regime., Comment: 13 pages, 5 figures

The study of quantum information processing seeks to characterize the resources that enable quantum information processing to perform tasks that are unfeasible or inefficient for classical information processing. Quantum cryptography is one such task, and researchers have identified entanglement as a sufficient resource for secure key generation. However, quantum discord, another type of quantum correlation beyond entanglement, has been found to be necessary for guaranteeing secure communication due to its direct relation to information leakage. Despite this, it is a long-standing problem how to make use of discord to analyze security in a specific quantum cryptography protocol. Here, based on our proposed quantum discord witness recently, we successfully address this issue by considering a BB84-like quantum key distribution protocol and its equivalent entanglement-based version. Our method is robust against imperfections in qubit sources and qubit measurements as well as basis misalignment due to quantum channels, which results in a better key rate than standard BB84 protocol. Those advantages are experimentally demonstrated via photonic phase encoding systems, which shows the practicality of our results.

Advanced Persistent Threat (APT) attack usually refers to the form of long-term, covert and sustained attack on specific targets, with an adversary using advanced attack techniques to destroy the key facilities of an organization. APT attacks have caused serious security threats and massive financial loss worldwide. Academics and industry thereby have proposed a series of solutions to detect APT attacks, such as dynamic/static code analysis, traffic detection, sandbox technology, endpoint detection and response (EDR), etc. However, existing defenses are failed to accurately and effectively defend against the current APT attacks that exhibit strong persistent, stealthy, diverse and dynamic characteristics due to the weak data source integrity, large data processing overhead and poor real-time performance in the process of real-world scenarios. To overcome these difficulties, in this paper we propose APTSHIELD, a stable, efficient and real-time APT detection system for Linux hosts. In the aspect of data collection, audit is selected to stably collect kernel data of the operating system so as to carry out a complete portrait of the attack based on comprehensive analysis and comparison of existing logging tools; In the aspect of data processing, redundant semantics skipping and non-viable node pruning are adopted to reduce the amount of data, so as to reduce the overhead of the detection system; In the aspect of attack detection, an APT attack detection framework based on ATT\&CK model is designed to carry out real-time attack response and alarm through the transfer and aggregation of labels. Experimental results on both laboratory and Darpa Engagement show that our system can effectively detect web vulnerability attacks, file-less attacks and remote access trojan attacks, and has a low false positive rate, which adds far more value than the existing frontier work.

A cyber-attack is a malicious attempt by experienced hackers to breach the target information system. Usually, the cyber-attacks are characterized as hybrid TTPs (Tactics, Techniques, and Procedures) and long-term adversarial behaviors, making the traditional intrusion detection methods ineffective. Most existing cyber-attack detection systems are implemented based on manually designed rules by referring to domain knowledge (e.g., threat models, threat intelligences). However, this process is lack of intelligence and generalization ability. Aiming at this limitation, this paper proposes an intelligent cyber-attack detection method based on provenance data. To effective and efficient detect cyber-attacks from a huge number of system events in the provenance data, we firstly model the provenance data by a heterogeneous graph to capture the rich context information of each system entities (e.g., process, file, socket, etc.), and learns a semantic vector representation for each system entity. Then, we perform online cyber-attack detection by sampling a small and compact local graph from the heterogeneous graph, and classifying the key system entities as malicious or benign. We conducted a series of experiments on two provenance datasets with real cyber-attacks. The experiment results show that the proposed method outperforms other learning based detection models, and has competitive performance against state-of-the-art rule based cyber-attack detection systems., Comment: 12pages,7figures,40 references

Untrusted node networks initially implemented by measurement-device-independent quantum key distribution (MDI-QKD) protocol are a crucial step on the roadmap of the quantum Internet. Considering extensive QKD implementations of trusted node networks, a workable upgrading tactic of existing networks toward MDI networks needs to be explicit. Here, referring to the nonstandalone (NSA) network of 5G, we propose an NSA-MDI scheme as an evolutionary selection for existing phase-encoding BB84 networks. Our solution can upgrade the BB84 networks and terminals that employ various phase-encoding schemes to immediately support MDI without hardware changes. This cost-effective upgrade effectively promotes the deployment of MDI networks as a step of untrusted node networks while taking full advantage of existing networks. In addition, the diversified demands on security and bandwidth are satisfied, and network survivability is improved.

Measurement-device-independent quantum key distribution (MDIQKD) is a revolutionary protocol since it is physically immune to all attacks on the detection side. However, the protocol still keeps the strict assumptions on the source side that the four BB84-states must be perfectly prepared to ensure security. Some protocols release part of the assumptions in the encoding system to keep the practical security, but the performance would be dramatically reduced. In this work, we present a MDIQKD protocol that requires less knowledge of encoding system to combat the troublesome modulation errors and fluctuations. We have also experimentally demonstrated the protocol. The result indicates the high-performance and good security for its practical applications. Besides, its robustness and flexibility exhibit a good value for complex scenarios such as the QKD networks., Comment: 22 pages, 9 figures