Search

Recognizing the promise of natural language interfaces to databases, prior studies have emphasized the development of text-to-SQL systems. While substantial progress has been made in this field, existing research has concentrated on generating SQL statements from text queries. The broader challenge, however, lies in inferring new information about the returned data. Our research makes two major contributions to address this gap. First, we introduce a novel Internet-of-Things (IoT) text-to-SQL dataset comprising 10,985 text-SQL pairs and 239,398 rows of network traffic activity. The dataset contains additional query types limited in prior text-to-SQL datasets, notably temporal-related queries. Our dataset is sourced from a smart building's IoT ecosystem exploring sensor read and network traffic data. Second, our dataset allows two-stage processing, where the returned data (network traffic) from a generated SQL can be categorized as malicious or not. Our results show that joint training to query and infer information about the data can improve overall text-to-SQL performance, nearly matching substantially larger models. We also show that current large language models (e.g., GPT3.5) struggle to infer new information about returned data, thus our dataset provides a novel test bed for integrating complex domain-specific reasoning into LLMs.

Recommender systems have been successfully applied in many applications. Nonetheless, recent studies demonstrate that recommender systems are vulnerable to membership inference attacks (MIAs), leading to the leakage of users' membership privacy. However, existing MIAs relying on shadow training suffer a large performance drop when the attacker lacks knowledge of the training data distribution and the model architecture of the target recommender system. To better understand the privacy risks of recommender systems, we propose shadow-free MIAs that directly leverage a user's recommendations for membership inference. Without shadow training, the proposed attack can conduct MIAs efficiently and effectively under a practice scenario where the attacker is given only black-box access to the target recommender system. The proposed attack leverages an intuition that the recommender system personalizes a user's recommendations if his historical interactions are used by it. Thus, an attacker can infer membership privacy by determining whether the recommendations are more similar to the interactions or the general popular items. We conduct extensive experiments on benchmark datasets across various recommender systems. Remarkably, our attack achieves far better attack accuracy with low false positive rates than baselines while with a much lower computational cost., Comment: This paper has been accepted by IJCAI-24

Recent booming development of Generative Artificial Intelligence (GenAI) has facilitated an emerging model commercialization for the purpose of reinforcement on model performance, such as licensing or trading Deep Neural Network (DNN) models. However, DNN model trading may trigger concerns of the unauthorized replications or misuses over the model, so that the benefit of the model ownership will be violated. Model identity auditing is a challenging issue in protecting intellectual property of DNN models and verifying the integrity and ownership of models for guaranteeing trusts in transactions is one of the critical obstacles. In this paper, we focus on the above issue and propose a novel Accumulator-enabled Auditing for Distributed Identity of DNN Model (A2-DIDM) that utilizes blockchain and zero-knowledge techniques to protect data and function privacy while ensuring the lightweight on-chain ownership verification. The proposed model presents a scheme of identity records via configuring model weight checkpoints with corresponding zero-knowledge proofs, which incorporates predicates to capture incremental state changes in model weight checkpoints. Our scheme ensures both computational integrity of DNN training process and programmability, so that the uniqueness of the weight checkpoint sequence in a DNN model is preserved, ensuring the correctness of the model identity auditing. In addition, A2-DIDM also addresses privacy protections in distributed identity via a proposed method of accumulators. We systematically analyze the security and robustness of our proposed model and further evaluate the effectiveness and usability of auditing DNN model identities.

A large number of federated learning (FL) algorithms have been proposed for different applications and from varying perspectives. However, the evaluation of such approaches often relies on a single metric (e.g., accuracy). Such a practice fails to account for the unique demands and diverse requirements of different use cases. Thus, how to comprehensively evaluate an FL algorithm and determine the most suitable candidate for a designated use case remains an open question. To mitigate this research gap, we introduce the Holistic Evaluation Metrics (HEM) for FL in this work. Specifically, we collectively focus on three primary use cases, which are Internet of Things (IoT), smart devices, and institutions. The evaluation metric encompasses various aspects including accuracy, convergence, computational efficiency, fairness, and personalization. We then assign a respective importance vector for each use case, reflecting their distinct performance requirements and priorities. The HEM index is finally generated by integrating these metric components with their respective importance vectors. Through evaluating different FL algorithms in these three prevalent use cases, our experimental results demonstrate that HEM can effectively assess and identify the FL algorithms best suited to particular scenarios. We anticipate this work sheds light on the evaluation process for pragmatic FL algorithms in real-world applications.

Adversarial generative models, such as Generative Adversarial Networks (GANs), are widely applied for generating various types of data, i.e., images, text, and audio. Accordingly, its promising performance has led to the GAN-based adversarial attack methods in the white-box and black-box attack scenarios. The importance of transferable black-box attacks lies in their ability to be effective across different models and settings, more closely aligning with real-world applications. However, it remains challenging to retain the performance in terms of transferable adversarial examples for such methods. Meanwhile, we observe that some enhanced gradient-based transferable adversarial attack algorithms require prolonged time for adversarial sample generation. Thus, in this work, we propose a novel algorithm named GE-AdvGAN to enhance the transferability of adversarial samples whilst improving the algorithm's efficiency. The main approach is via optimising the training process of the generator parameters. With the functional and characteristic similarity analysis, we introduce a novel gradient editing (GE) mechanism and verify its feasibility in generating transferable samples on various models. Moreover, by exploring the frequency domain information to determine the gradient editing direction, GE-AdvGAN can generate highly transferable adversarial samples while minimizing the execution time in comparison to the state-of-the-art transferable adversarial attack algorithms. The performance of GE-AdvGAN is comprehensively evaluated by large-scale experiments on different datasets, which results demonstrate the superiority of our algorithm. The code for our algorithm is available at: https://github.com/LMBTough/GE-advGAN, Comment: Accepted by SIAM International Conference on Data Mining (SDM24)

As artificial intelligence (AI) increasingly becomes an integral part of our societal and individual activities, there is a growing imperative to develop responsible AI solutions. Despite a diverse assortment of machine learning fairness solutions is proposed in the literature, there is reportedly a lack of practical implementation of these tools in real-world applications. Industry experts have participated in thorough discussions on the challenges associated with operationalising fairness in the development of machine learning-empowered solutions, in which a shift toward human-centred approaches is promptly advocated to mitigate the limitations of existing techniques. In this work, we propose a human-in-the-loop approach for fairness auditing, presenting a mixed visual analytical system (hereafter referred to as 'FairCompass'), which integrates both subgroup discovery technique and the decision tree-based schema for end users. Moreover, we innovatively integrate an Exploration, Guidance and Informed Analysis loop, to facilitate the use of the Knowledge Generation Model for Visual Analytics in FairCompass. We evaluate the effectiveness of FairCompass for fairness auditing in a real-world scenario, and the findings demonstrate the system's potential for real-world deployability. We anticipate this work will address the current gaps in research for fairness and facilitate the operationalisation of fairness in machine learning systems., Comment: Accepted in IEEE Transactions on Artificial Intelligence

To better understand the output of deep neural networks (DNN), attribution based methods have been an important approach for model interpretability, which assign a score for each input dimension to indicate its importance towards the model outcome. Notably, the attribution methods use the axioms of sensitivity and implementation invariance to ensure the validity and reliability of attribution results. Yet, the existing attribution methods present challenges for effective interpretation and efficient computation. In this work, we introduce MFABA, an attribution algorithm that adheres to axioms, as a novel method for interpreting DNN. Additionally, we provide the theoretical proof and in-depth analysis for MFABA algorithm, and conduct a large scale experiment. The results demonstrate its superiority by achieving over 101.5142 times faster speed than the state-of-the-art attribution algorithms. The effectiveness of MFABA is thoroughly evaluated through the statistical analysis in comparison to other methods, and the full implementation package is open-source at: https://github.com/LMBTough/MFABA, Comment: Accepted by The 38th Annual AAAI Conference on Artificial Intelligence (AAAI-24)

Vertical federated learning has garnered significant attention as it allows clients to train machine learning models collaboratively without sharing local data, which protects the client's local private data. However, existing VFL methods face challenges when dealing with heterogeneous local models among participants, which affects optimization convergence and generalization. To address this challenge, this paper proposes a novel approach called Vertical federated learning for training multiple Heterogeneous models (VFedMH). VFedMH focuses on aggregating the local embeddings of each participant's knowledge during forward propagation. To protect the participants' local embedding values, we propose an embedding protection method based on lightweight blinding factors. In particular, participants obtain local embedding using local heterogeneous models. Then the passive party, who owns only features of the sample, injects the blinding factor into the local embedding and sends it to the active party. The active party aggregates local embeddings to obtain global knowledge embeddings and sends them to passive parties. The passive parties then utilize the global embeddings to propagate forward on their local heterogeneous networks. However, the passive party does not own the sample labels, so the local model gradient cannot be calculated locally. To overcome this limitation, the active party assists the passive party in computing its local heterogeneous model gradients. Then, each participant trains their local model using the heterogeneous model gradients. The objective is to minimize the loss value of their respective local heterogeneous models. Extensive experiments are conducted to demonstrate that VFedMH can simultaneously train multiple heterogeneous models with heterogeneous optimization and outperform some recent methods in model performance.

Federated learning (FL) is a popular approach to facilitate privacy-aware machine learning since it allows multiple clients to collaboratively train a global model without granting others access to their private data. It is, however, known that FL can be vulnerable to membership inference attacks (MIAs), where the training records of the global model can be distinguished from the testing records. Surprisingly, research focusing on the investigation of the source inference problem appears to be lacking. We also observe that identifying a training record's source client can result in privacy breaches extending beyond MIAs. For example, consider an FL application where multiple hospitals jointly train a COVID-19 diagnosis model, membership inference attackers can identify the medical records that have been used for training, and any additional identification of the source hospital can result the patient from the particular hospital more prone to discrimination. Seeking to contribute to the literature gap, we take the first step to investigate source privacy in FL. Specifically, we propose a new inference attack (hereafter referred to as source inference attack -- SIA), designed to facilitate an honest-but-curious server to identify the training record's source client. The proposed SIAs leverage the Bayesian theorem to allow the server to implement the attack in a non-intrusive manner without deviating from the defined FL protocol. We then evaluate SIAs in three different FL frameworks to show that in existing FL frameworks, the clients sharing gradients, model parameters, or predictions on a public dataset will leak such source information to the server. We also conduct extensive experiments on various datasets to investigate the key factors in an SIA. The experimental results validate the efficacy of the proposed SIAs., Comment: Accepted by IEEE Transactions on Dependable and Secure Computing

This paper examines the ethical considerations and implications of large language models (LLMs) in generating content. It highlights the potential for both positive and negative uses of generative AI programs and explores the challenges in assigning responsibility for their outputs. The discussion emphasizes the need for proactive ethical frameworks and policy measures to guide the responsible development and deployment of LLMs., Comment: 5 pages

The ongoing deployment of the fifth generation (5G) wireless networks constantly reveals limitations concerning its original concept as a key driver of Internet of Everything (IoE) applications. These 5G challenges are behind worldwide efforts to enable future networks, such as sixth generation (6G) networks, to efficiently support sophisticated applications ranging from autonomous driving capabilities to the Metaverse. Edge learning is a new and powerful approach to training models across distributed clients while protecting the privacy of their data. This approach is expected to be embedded within future network infrastructures, including 6G, to solve challenging problems such as resource management and behavior prediction. This survey article provides a holistic review of the most recent research focused on edge learning vulnerabilities and defenses for 6G-enabled IoT. We summarize the existing surveys on machine learning for 6G IoT security and machine learning-associated threats in three different learning modes: centralized, federated, and distributed. Then, we provide an overview of enabling emerging technologies for 6G IoT intelligence. Moreover, we provide a holistic survey of existing research on attacks against machine learning and classify threat models into eight categories, including backdoor attacks, adversarial examples, combined attacks, poisoning attacks, Sybil attacks, byzantine attacks, inference attacks, and dropping attacks. In addition, we provide a comprehensive and detailed taxonomy and a side-by-side comparison of the state-of-the-art defense methods against edge learning vulnerabilities. Finally, as new attacks and defense technologies are realized, new research and future overall prospects for 6G-enabled IoT are discussed., Comment: This paper has been accepted for publication in IEEE Communications Surveys \& Tutorials

Electronic medical records (EMRs) are stored in relational databases. It can be challenging to access the required information if the user is unfamiliar with the database schema or general database fundamentals. Hence, researchers have explored text-to-SQL generation methods that provide healthcare professionals direct access to EMR data without needing a database expert. However, currently available datasets have been essentially "solved" with state-of-the-art models achieving accuracy greater than or near 90%. In this paper, we show that there is still a long way to go before solving text-to-SQL generation in the medical domain. To show this, we create new splits of the existing medical text-to-SQL dataset MIMICSQL that better measure the generalizability of the resulting models. We evaluate state-of-the-art language models on our new split showing substantial drops in performance with accuracy dropping from up to 92% to 28%, thus showing substantial room for improvement. Moreover, we introduce a novel data augmentation approach to improve the generalizability of the language models. Overall, this paper is the first step towards developing more robust text-to-SQL models in the medical domain.\footnote{The dataset and code will be released upon acceptance.

Federated edge learning can be essential in supporting privacy-preserving, artificial intelligence (AI)-enabled activities in digital twin 6G-enabled Internet of Things (IoT) environments. However, we need to also consider the potential of attacks targeting the underlying AI systems (e.g., adversaries seek to corrupt data on the IoT devices during local updates or corrupt the model updates); hence, in this article, we propose an anticipatory study for poisoning attacks in federated edge learning for digital twin 6G-enabled IoT environments. Specifically, we study the influence of adversaries on the training and development of federated learning models in digital twin 6G-enabled IoT environments. We demonstrate that attackers can carry out poisoning attacks in two different learning settings, namely: centralized learning and federated learning, and successful attacks can severely reduce the model's accuracy. We comprehensively evaluate the attacks on a new cyber security dataset designed for IoT applications with three deep neural networks under the non-independent and identically distributed (Non-IID) data and the independent and identically distributed (IID) data. The poisoning attacks, on an attack classification problem, can lead to a decrease in accuracy from 94.93% to 85.98% with IID data and from 94.18% to 30.04% with Non-IID., Comment: The paper is accepted and will be published in the IEEE ICC 2023 Conference Proceedings

Underwater drones have found a place in oceanography, oceanic research, bathymetric surveys, military, surveillance, monitoring, undersea exploration, mining, commercial diving, photography and several other activities. Drones housed with several sensors and complex propulsion systems help oceanographic scientists and undersea explorers to map the seabed, study waves, view dead zones, analyze fish counts, predict tidal wave behaviors, aid in finding shipwrecks, building windfarms, examine oil platforms located in deep seas and inspect nuclear reactors in the ship vessels. While drones can be explicitly programmed for specific missions, data security and privacy are crucial issues of serious concern. Blockchain has emerged as a key enabling technology, amongst other disruptive technological enablers, to address security, data sharing, storage, process tracking, collaboration and resource management. This study presents a comprehensive review on the utilization of Blockchain in different underwater applications, discussing use cases and detailing benefits. Potential challenges of underwater applications addressed by Blockchain have been detailed. This work identifies knowledge gaps between theoretical research and real-time Blockchain integration in realistic underwater drone applications. The key limitations for effective integration of Blockchain in real-time integration in UUD applications, along with directions for future research have been presented.

Network Virtualization (NV) is an emerging network dynamic planning technique to overcome network rigidity. As its necessary challenge, Virtual Network Embedding (VNE) enhances the scalability and flexibility of the network by decoupling the resources and services of the underlying physical network. For the future multi-domain physical network modeling with the characteristics of dynamics, heterogeneity, privacy, and real-time, the existing related works perform satisfactorily. Federated learning (FL) jointly optimizes the network by sharing parameters among multiple parties and is widely employed in disputes over data privacy and data silos. Aiming at the NV challenge of multi-domain physical networks, this work is the first to propose using FL to model VNE, and presents a VNE architecture based on Horizontal Federated Learning (HFL) (HFL-VNE). Specifically, combined with the distributed training paradigm of FL, we deploy local servers in each physical domain, which can effectively focus on local features and reduce resource fragmentation. A global server is deployed to aggregate and share training parameters, which enhances local data privacy and significantly improves learning efficiency. Furthermore, we deploy the Deep Reinforcement Learning (DRL) model in each server to dynamically adjust and optimize the resource allocation of the multi-domain physical network. In DRL-assisted FL, HFL-VNE jointly optimizes decision-making through specific local and federated reward mechanisms and loss functions. Finally, the superiority of HFL-VNE is proved by combining simulation experiments and comparing it with related works.

Cryptocurrency has been (ab)used to purchase illicit goods and services such as drugs, weapons and child pornography (also referred to as child sexual abuse materials), and thus mobile devices (where cryptocurrency wallet applications are installed) are a potential source of evidence in a criminal investigation. Not surprisingly, there has been increased focus on the security of cryptocurrency wallets, although forensic extraction and attribution of forensic artefacts from such wallets is understudied. In this paper, we examine Bitcoin and Dogecoin. The latter is increasingly popular partly due to endorsements from celebrities and being positioned as an introductory path to cryptocurrency for newcomers. Specifically, we demonstrate how one can acquire forensic artefacts from Android Bitcoin and Dogecoin cryptocurrency wallets, such as wallet IDs, transaction IDs, timestamp information, email addresses, cookies, and OAuth tokens.

Cryptocurrencies are no longer just the preferred option for cybercriminal activities on darknets, due to the increasing adoption in mainstream applications. This is partly due to the transparency associated with the underpinning ledgers, where any individual can access the record of a transaction record on the public ledger. In this paper, we build a dataset comprising Bitcoin transactions between 12 July 2019 and 26 May 2021. This dataset (hereafter referred to as BABD-13) contains 13 types of Bitcoin addresses, 5 categories of indicators with 148 features, and 544,462 labeled data, which is the largest labeled Bitcoin address behavior dataset publicly available to our knowledge. We then use our proposed dataset on common machine learning models, namely: k-nearest neighbors algorithm, decision tree, random forest, multilayer perceptron, and XGBoost. The results show that the accuracy rates of these machine learning models for the multi-classification task on our proposed dataset are between 93.24% and 97.13%. We also analyze the proposed features and their relationships from the experiments, and propose a k-hop subgraph generation algorithm to extract a k-hop subgraph from the entire Bitcoin transaction graph constructed by the directed heterogeneous multigraph starting from a specific Bitcoin address node (e.g., a known transaction associated with a criminal investigation). Besides, we initially analyze the behavior patterns of different types of Bitcoin addresses according to the extracted features., Comment: 14 pages, 4 figures

Advances in synthetic biology and nanotechnology have contributed to the design of tools that can be used to control, reuse, modify, and re-engineer cells' structure, as well as enabling engineers to effectively use biological cells as programmable substrates to realize Bio-Nano Things (biological embedded computing devices). Bio-NanoThings are generally tiny, non-intrusive, and concealable devices that can be used for in-vivo applications such as intra-body sensing and actuation networks, where the use of artificial devices can be detrimental. Such (nano-scale) devices can be used in various healthcare settings such as continuous health monitoring, targeted drug delivery, and nano-surgeries. These services can also be grouped to form a collaborative network (i.e., nanonetwork), whose performance can potentially be improved when connected to higher bandwidth external networks such as the Internet, say via 5G. However, to realize the IoBNT paradigm, it is also important to seamlessly connect the biological environment with the technological landscape by having a dynamic interface design to convert biochemical signals from the human body into an equivalent electromagnetic signal (and vice versa). This, unfortunately, risks the exposure of internal biological mechanisms to cyber-based sensing and medical actuation, with potential security and privacy implications. This paper comprehensively reviews bio-cyber interface for IoBNT architecture, focusing on bio-cyber interfacing options for IoBNT like biologically inspired bio-electronic devices, RFID enabled implantable chips, and electronic tattoos. This study also identifies known and potential security and privacy vulnerabilities and mitigation strategies for consideration in future IoBNT designs and implementations., Comment: 41 pages, 9 tables, 6 figures

To ensure the privacy of sensitive data used in the training of deep learning models, a number of privacy-preserving methods have been designed by the research community. However, existing schemes are generally designed to work with textual data, or are not efficient when a large number of images is used for training. Hence, in this paper we propose a lightweight and efficient approach to preserve image privacy while maintaining the availability of the training set. Specifically, we design the pixel block mixing algorithm for image classification privacy preservation in deep learning. To evaluate its utility, we use the mixed training set to train the ResNet50, VGG16, InceptionV3 and DenseNet121 models on the WIKI dataset and the CNBC face dataset. Experimental findings on the testing set show that our scheme preserves image privacy while maintaining the availability of the training set in the deep learning models. Additionally, the experimental results demonstrate that we achieve good performance for the VGG16 model on the WIKI dataset and both ResNet50 and DenseNet121 on the CNBC dataset. The pixel block algorithm achieves fairly high efficiency in the mixing of the images, and it is computationally challenging for the attackers to restore the mixed training set to the original training set. Moreover, data augmentation can be applied to the mixed training set to improve the training's effectiveness., Comment: 11 pages, 16 figures

As smartphones and smartphone applications are widely used in a healthcare context (e.g., remote healthcare), these devices and applications may need to comply with the Health Insurance Portability and Accountability Act (HIPAA) of 1996. In other words, adequate safeguards to protect the user's sensitive information (e.g., personally identifiable information and/or medical history) are required to be enforced on such devices and applications. In this study, we forensically focus on the potential of recovering residual data from Android medical applications, with the objective of providing an initial risk assessment of such applications. Our findings (e.g., documentation of the artifacts) also contribute to a better understanding of the types and location of evidential artifacts that can, potentially, be recovered from these applications in a digital forensic investigation., Comment: Presented at the 5th EAI International Conference on Future Access Enablers of Ubiquitous and Intelligent Infrastructures (EAI FABULOUS 2021), Zagreb, Croatia

Internet of Things (IoT) devices are becoming ubiquitous in our lives, with applications spanning from the consumer domain to commercial and industrial systems. The steep growth and vast adoption of IoT devices reinforce the importance of sound and robust cybersecurity practices during the device development life-cycles. IoT-related vulnerabilities, if successfully exploited can affect, not only the device itself, but also the application field in which the IoT device operates. Evidently, identifying and addressing every single vulnerability is an arduous, if not impossible, task. Attack taxonomies can assist in classifying attacks and their corresponding vulnerabilities. Security countermeasures and best practices can then be leveraged to mitigate threats and vulnerabilities before they emerge into catastrophic attacks and ensure overall secure IoT operation. Therefore, in this paper, we provide an attack taxonomy which takes into consideration the different layers of IoT stack, i.e., device, infrastructure, communication, and service, and each layer's designated characteristics which can be exploited by adversaries. Furthermore, using nine real-world cybersecurity incidents, that had targeted IoT devices deployed in the consumer, commercial, and industrial sectors, we describe the IoT-related vulnerabilities, exploitation procedures, attacks, impacts, and potential mitigation mechanisms and protection strategies. These (and many other) incidents highlight the underlying security concerns of IoT systems and demonstrate the potential attack impacts of such connected ecosystems, while the proposed taxonomy provides a systematic procedure to categorize attacks based on the affected layer and corresponding impact., Comment: IEEE Internet of Things Journal

The need for medical image encryption is increasingly pronounced, for example to safeguard the privacy of the patients' medical imaging data. In this paper, a novel deep learning-based key generation network (DeepKeyGen) is proposed as a stream cipher generator to generate the private key, which can then be used for encrypting and decrypting of medical images. In DeepKeyGen, the generative adversarial network (GAN) is adopted as the learning network to generate the private key. Furthermore, the transformation domain (that represents the "style" of the private key to be generated) is designed to guide the learning network to realize the private key generation process. The goal of DeepKeyGen is to learn the mapping relationship of how to transfer the initial image to the private key. We evaluate DeepKeyGen using three datasets, namely: the Montgomery County chest X-ray dataset, the Ultrasonic Brachial Plexus dataset, and the BraTS18 dataset. The evaluation findings and security analysis show that the proposed key generation network can achieve a high-level security in generating the private key.

Image forensic plays a crucial role in both criminal investigations (e.g., dissemination of fake images to spread racial hate or false narratives about specific ethnicity groups) and civil litigation (e.g., defamation). Increasingly, machine learning approaches are also utilized in image forensics. However, there are also a number of limitations and vulnerabilities associated with machine learning-based approaches, for example how to detect adversarial (image) examples, with real-world consequences (e.g., inadmissible evidence, or wrongful conviction). Therefore, with a focus on image forensics, this paper surveys techniques that can be used to enhance the robustness of machine learning-based binary manipulation detectors in various adversarial scenarios., Comment: 37 pages, 24 figures, Accepted to the Journal Computer and Security (Elsevier)

Federated learning has become prevalent in medical diagnosis due to its effectiveness in training a federated model among multiple health institutions (i.e. Data Islands (DIs)). However, increasingly massive DI-level poisoning attacks have shed light on a vulnerability in federated learning, which inject poisoned data into certain DIs to corrupt the availability of the federated model. Previous works on federated learning have been inadequate in ensuring the privacy of DIs and the availability of the final federated model. In this paper, we design a secure federated learning mechanism with multiple keys to prevent DI-level poisoning attacks for medical diagnosis, called SFPA. Concretely, SFPA provides privacy-preserving random forest-based federated learning by using the multi-key secure computation, which guarantees the confidentiality of DI-related information. Meanwhile, a secure defense strategy over encrypted locally-submitted models is proposed to defense DI-level poisoning attacks. Finally, our formal security analysis and empirical tests on a public cloud platform demonstrate the security and efficiency of SFPA as well as its capability of resisting DI-level poisoning attacks.