Abstract Sumplete is a logic puzzle generated by ChatGPT in March 2023. The puzzle consists of a rectangular grid, with each cell containing an integer. Each row and column also has an integer called target value assigned to it. The objective of this puzzle is to cross out some numbers in the grid such that the sum of uncrossed numbers in each row and column is equal to the corresponding target value. In this paper, we prove that Sumplete is NP-complete. We also propose a physical zero-knowledge proof protocol for the puzzle using physical cards.
Chenfei Nie, Yannan Li, Yuxin Yang, Yuede Ji · 5 authors
Federated learning (FL) is an emerging distributed learning paradigm without sharing participating clients' private data. However, existing works show that FL is vulnerable to both Byzantine (security) attacks and data reconstruction (privacy) attacks. Almost all the existing FL defenses only address one of the two attacks. A few defenses address the two attacks, but they are not efficient and effective enough. We propose BPFL, an efficient Byzantine-robust and provably privacy-preserving FL method that addresses all the issues. Specifically, we draw on state-of-the-art Byzantine-robust FL methods and use similarity metrics to measure the robustness of each participating client in FL. The validity of clients are formulated as circuit constraints on similarity metrics and verified via a zero-knowledge proof. Moreover, the client models are masked by a shared random vector, which is generated based on homomorphic encryption. In doing so, the server receives the masked client models rather than the true ones, which are proven to be private. BPFL is also efficient due to the usage of non-interactive zero-knowledge proof. Experimental results on various datasets show that our BPFL is efficient, Byzantine-robust, and privacy-preserving.
This study proposes an privacy-protection method for blockchain transactions based on lightweight homomorphic encryption, aiming to ensure the security of transaction data and user privacy, and improve transaction efficiency. We have built a blockchain infrastructure and, based on its structural characteristics, adopted zero-knowledge proof technology to verify the legitimacy of data, ensuring the authenticity and accuracy of transactions from the application end to the smart-contract end. On this basis, the Paillier algorithm is used for key generation, encryption, and decryption, and intelligent protection of blockchain transaction privacy is achieved through a secondary encryption mechanism. The experimental results show that this method performs well in privacy and security protection, with a data leakage probability as low as 2.8%, and can effectively defend against replay attacks and forged-transaction attacks. The degree of confusion remains above 0.9, with small fluctuations and short running time under different key lengths and moderate CPU usage, achieving lightweight homomorphic encryption. This not only ensures the security and privacy of transaction data in blockchain networks, but also reduces computational complexity and resource consumption, better adapting to the high-concurrency and low-latency characteristics of blockchain networks, thereby ensuring the efficiency and real-time performance of transactions.
Change Institutions to: University of Waterloo, Yue Zhao, Claudio Angione, Harry Yang · 8 authors
The rapid advancement of ML models in critical sectors such as healthcare, finance, and security has intensified the need for robust data security, model integrity, and reliable outputs. Large multimodal foundational models, while crucial for complex tasks, present challenges in scalability, reliability, and potential misuse. Decentralized systems offer a solution by distributing workload and mitigating central points of failure, but they introduce risks of unauthorized access to sensitive data across nodes. We address these challenges with a comprehensive framework designed for responsible AI development. Our approach incorporates: 1) Zero-knowledge proofs for secure model verification, enhancing trust without compromising privacy. 2) Consensus-based verification checks to ensure consistent outputs across nodes, mitigating hallucinations and maintaining model integrity. 3) Split Learning techniques that segment models across different nodes, preserving data privacy by preventing full data access at any point. 4) Hardware-based security through trusted execution environments (TEEs) to protect data and computations. This framework aims to enhance security and privacy and improve the reliability and fairness of multimodal AI systems. Promoting efficient resource utilization contributes to more sustainable AI development. Our state-of-the-art proofs and principles demonstrate the framework's effectiveness in responsibly democratizing artificial intelligence, offering a promising approach for building secure and private foundational models.
Federated Learning (FL) is a technique in the field of machine learning that prioritizes privacy by allowing collaborative model training without revealing data. This article explores the basics of FL and its importance in protecting data privacy in sectors such as healthcare, finance, and industrial engineering. By using data sources FL enables the development of strong and adaptable AI models without centralizing sensitive information. We delve into the methodologies behind FL including secure multiparty computation, differential privacy, and homomorphic encryption. Additionally, we look at the ways FL is used, such as speeding up medical research improving financial security and streamlining industrial processes. The challenges related to FL - like communication diverse data distributions and scalability - are also addressed. Lastly, we discuss trends, in FL that focus on enhancing privacy techniques and complying with regulations. This thorough overview highlights how FL can revolutionize AI advancement while upholding privacy standards. Keywords: Federated Learning, Privacy Preservation, Decentralized Machine Learning, Secure Multiparty Computation, Differential Privacy, Healthcare AI, Industrial Engineering, Data Silos, Collaborative Learning.
Togzhan Barakbayeva, Zhuo Cai, Amir Kafshdar Goharshady, Karaneh Keypoor
Correlated equilibria are a standard solution concept in game theory and generalize Nash equilibria. In a 2-player non-cooperative game in which player i has action set A_i, a correlated equilibrium is a self-enforcing probability distribution σ over A_1 * A_2. Specifically, when a strategy profile (s_1, s_2) in A_1 * A_2 is sampled according to σ, each player i can observe their own component s_i, but not the other player's component. Knowing s_i and σ, player i cannot increase their expected payoff by defecting and playing a strategy s'_i different from s_i. Correlated equilibria are ubiquitous and crucial in mechanism design, including in the design of blockchain-based protocols which aim to incentivize honest behavior. A correlated equilibrium depends on a centralized and impartial oracle, often called the ''external signal'' in game theory literature, to sample a strategy profile and disclose each player's component to them, while keeping the other player's component secret. However, there is currently no trustless method to achieve this on the blockchain without centralization or relying on trusted third-parties. In this work, we address this challenge and provide two novel protocols, one based on oblivious transfer and the other based on zkSNARKs to replace the public signal with a smart contract. We prove that our approaches are secure and provide the desired privacy properties of a correlated equilibrium, while also being efficient in terms of gas usage and thus affordable in practice.
Blockchain is a promising infrastructure for the internet and digital economy, but it has serious scalability problems, that is, long block synchronization time and high storage cost. Conventional coarse-grained data deduplication schemes (block or file level) are proved to be ineffective on improving the scalability of blockchains. Based on comprehensive analysis on typical blockchain workloads, we propose two new locality concepts (economic and argument locality) and a novel fine-grained data deduplication scheme (transaction level) named Alias-Chain. Specifically, Alias-Chain replaces frequently used data, for example, smart contract arguments, with much shorter aliases to reduce the block sizes, which results in both shorter synchronization time and lower storage cost. Furthermore, to solve the potential consistency issue in Alias-Chain, we propose two complementary techniques: one is generating aliases from history blocks with high consistency, and the other is speeding up the generation of aliases via a specific algorithm. Our simulation results show: (1) the average transfer and SC-call transaction (a transaction used to call the smart contracts in the blockchain) sizes can be significantly reduced by up to 11.03% and 79.44% in native Ethereum, and up to 39.29% and 81.84% in Ethereum optimized by state-of-the-art techniques; and (2) the two complementary techniques well address the inconsistency risk with very limited impact on the benefit of Alias-Chain. Prototyping-based experiments are further conducted on a testbed consisting of up to 3200 miners. The results demonstrate the effectiveness and efficiency of Alias-Chain on reducing block synchronization time and storage cost under typical real-world workloads.
In order to protect the cooperative training of machine learning models across decentralized devices, this study presents a novel privacy-preserving federated learning system with homomorphic encryption (PFed-HE). Starting with the careful selection and improvement of homomorphic encryption algorithms for effective and safe computations, the methodology takes a multifaceted approach. A unique approach for encrypting and aggregating gradients while maintaining privacy is introduced, utilizing homomorphic encryption and incorporating differential privacy techniques for an additional layer of confidentiality. The federated learning architecture includes a client-side encryption module, which smoothly integrates the encryption process into the model training workflow. A decentralized model aggregation approach allows encrypted model updates from numerous clients to be securely integrated while maintaining individual data privacy. To improve security, the methodology incorporates a dynamic key management system with periodic key rotation and a secure key agreement protocol for establishing shared encryption keys. Batch processing, parallelizing homomorphic encryption processes, and communication compression are the main tactics used in performance optimization to reduce computational overhead and improve scalability. The PFed-HE system is integrated with prominent federated learning frameworks, thoroughly evaluated in simulated real-world scenarios, and applied to specific use cases in fields such as healthcare and finance. This PFed-HE system tackles the ethical issues and openness that are critical for deployment in sensitive applications, in addition to showcasing advances in privacy-preserving machine learning.
In the rapidly evolving digital world, blockchain technology is becoming the foundation for numerous applications, ranging from financial services to supply chain management. As the usage of blockchain is becoming more prevalent, the energy-intensive nature of this technology has raised concerns about its long-term sustainability and environmental footprint. To address this challenge, we explore the potential of Peer-to-Peer Federated Learning (P2P-FL), a distributed machine learning approach that allows multiple nodes to collaborate without sharing raw data. We present a novel integration of P2P-FL with blockchain technology, aimed at enhancing the sustainability and efficiency of blockchain networks. The basic idea of our approach is the use of distributed learning mechanisms to find the optimal performance parameters of blockchain without relying on centralized control. These parameters are then used by a load-balancing mechanism that prioritizes energy efficiency to distribute loads on different blockchains. Furthermore, we formulate a non-cooperative game theory model to align the individual node strategies with the collective objective of energy optimization, ensuring a balance between self-interest and overall network performance. Our work is exemplified through a case study in the renewable energy sector, demonstrating the application of our model in creating an efficient marketplace for energy trading. The experimentation and results indicate a significant improvement in the execution times and energy consumption of blockchain networks. Therefore, the overall sustainability of the network is enhanced, making our framework practical and applicable in real-world scenarios.
Jingcheng Zhang, Yingxuan Ren, Man Ho Au, Ka-Ho Chow · 9 authors
Abstract With the rapid developments in sequencing technologies, individuals now have unprecedented access to their genomic data. However, existing data management systems or protocols are inadequate for protecting privacy, limiting individuals’ control over their genomic information, hindering data sharing, and posing a challenge for biomedical research. To fill the gap, an owner-governed system that fulfills owner authority, lifecycle data encryption, and verifiability at the same time is prompted. In this paper, we realized Governome, an owner-governed data management system designed to empower individuals with absolute control over their genomic data during data sharing. Governome uses a blockchain to manage all transactions and permissions, enabling data owners with dynamic permission management and to be fully informed about every data usage. It uses homomorphic encryption and zero-knowledge proofs to enable genomic data storage and computation in an encrypted and verifiable form for its whole lifecycle. Governome supports genomic analysis tasks, including individual variant query, cohort study, GWAS analysis, and forensics. Query of a variant’s genotype distribution among 2,504 1kGP individuals in Governome can be efficiently completed in under 18 hours on an ordinary server. Governome is an open-source project available at https://github.com/HKU-BAL/Governome .
Zero Knowledge Proof (ZKP) is a very effective method of preserving privacy as it hides the most confidential information throughout the transaction. In this paper, we present a security and privacy-preserving approach for blockchain that relies on account and multi-data asset models using the Zero Knowledge Proof (ZKP) mechanism. We provide options for transferring data assets and detecting duplicate expenditures, and we also develop transaction structures, anonymised addresses and anonymised metadata for the data assets. To create and validate the ZKP, we use the zk-SNARKs algorithm and specify validation criteria for masked transactions, and finally conduct experimental tests to validate it. Creating better algorithms for ZKP will be the focus of our future efforts.
Jianrong Wang, Yang Shi, Dengcheng Hu, Keqiu Li · 5 authors
One notable drawback of traditional Federated Learning (FL) is its susceptibility to single point of failures. In recent years, Blockchain-based Federated Learning (BCFL) has been proposed as an effective solution to address this issue. However, existing BCFL frameworks face challenges in heterogeneous IoT scenarios. The heterogeneity of IoT devices poses challenges to the adaptation of blockchain consensus. The integration of blockchain imposes constraints on the learning scalability of systems, making it challenging to accommodate a large number of heterogeneous IoT devices. On the other hand, current blockchain consensus fail to sufficiently measure the contributions and destructions among heterogeneous devices in terms of learning quality, leading to low learning security and insufficient incentive fairness. To overcome the limitations of prior art, this paper introduces CoCFL, a novel blockchain-based federated learning framework based on multi-chain collaborative model. CoCFL enhances learning scalability by adopting a multi-chain asynchronous collaboration approach that partitions both learning and communication granularity of the system. Within each sub chain, CoCFL introduces a lightweight, secure and incentive-fair blockchain-based federated learning consensus, called Proof of Contribution to FL (PoCFL). In PoCFL, partic-ipants' contributions to the learning and the consensus process form the basis for delegating consensus responsibility and dis-tributing rewards. Furthermore, we introduce a novel malicious model detection algorithm into PoCFL, called the Trustee Nearest Algorithm. Through Trustee Nearest, PoCFL effectively mitigates poisoning attacks. Experimental results demonstrate that CoCFL exhibits better learning scalability compared to traditional FL and and avdanced BCFL frameworks in the same scenarios and can effectively withstand poisoning attacks initiated by at least 40% of malicious participants. Moreover, CoCFL demonstrated good incentive fairness during the learning process.
S. Murali, S. Kanaga Suba Raja, E. Kanimozhi, D. Akila · 6 authors
In recent years, concerns about the security of data exchanged between the Internet of Things (IoT) devices have increased with respect to information security. Data security and the prevention of critical information leaks are both facilitated by a security model. Using Blockchain technology, the issues with data security in IoT devices may be resolved. Decentralized ledger technology known as blockchain enables many useful applications, including distributed storage, consensus, encryption, and Machine-to-Machine (M2M) data transmission. This paper develops a cross-layer framework for protecting the sensitive data of IoT by applying Blockchain based Federated Learning model (BCFL). The data records are organized into blocks in the BCFL paradigm, and then encode and decode methods are used to send the data to the servers. Using the Federated learning method, the block size and coding redundancy factor are calculated in a block-by-block transmission manner. To protect the integrity of data, a hash function and a digital signature are computed for each block of data. By experimental results, it was shown that the proposed BCFL model attains higher detection accuracy and correctness of data with reduced encoding/decoding time and reconstruction time.
In the period of decentralized networks and Hadoop technology, the confluence of Blockchain and Machine Learning (ML) technologies has emerged as a promising solution for ensuring robust, transparent, and private transactions. While the individual merits of Blockchain in ensuring data integrity and ML in deriving insights are well-established, their synergistic effects particularly in the realm of privacy preservation are yet to be fully explored. This fusion has the potential to revolutionize sectors like healthcare, finance, and supply chain by offering unprecedented levels of data privacy without compromising on system performance. This paper presents a comprehensive review of existing models that employ machine learning techniques for privacy preservation operations within blockchain frameworks. Utilizing a set of predetermined metrics delay, throughput, energy efficiency, privacy levels, deployment cost, and scalability the paper compares these models to provide a nuanced understanding of their capabilities and limitations. Our in-depth comparison elucidates the tradeoffs involved in selecting specific blockchain-ML models for diverse applications. For instance, while some models may offer higher throughput, they could compromise on privacy levels. Conversely, models excelling in privacy preservation might incur higher energy costs or delays. This multi-faceted evaluation not only guides stakeholders to choose the most efficient model based on individual or multiple metrics but also uncovers avenues for future research. By correlating these metrics with real-world use cases in healthcare, finance, and supply chain, among others, the paper adds a layer of practical applicability.
Federated learning (FL), a decentralized machine learning approach, enables privacy-preserving and collaborative model training without centralizing sensitive data. It has been successfully applied in various domains, including e-commerce, healthcare, and finance. However, existing FL schemes often fail to address personalized task requirements, such as prior-itizing the accuracy of specific classes within a dataset. The recent surge in Artificial Intelligence Generated Content (AIGC) offers potential to meet these personalized requirements by augmenting the training data of specific classes with generative models. Nevertheless, integrating generative models with FL introduces challenges, such as non-compliant data, disorganized distributions, and limited computing power on edge devices. To address these challenges, we propose AIGC-augmented Federated Preference Learning (FPL), which focuses on training specific data classes, referred to as preference classes (PCs). To improve the quality of AI -generated data, we implement strategies such as pre-training and fine-tuning across various datasets. Additionally, we enhance FL efficiency through a client selection strategy that matches generated data tasks with suitable clients and an AIGC data distribution strategy that optimally allocates data where it is most needed. We validate the feasibility and effectiveness of AIGC-augmented FPL by conducting experiments on the MNIST and CIFAR-10 datasets from various perspectives.
Abstract In response to the dual privacy protection challenges concerning the confidentiality of transaction amounts and identities in cross‐border trade, a transaction scheme that combines + HomEIG Zero Knowledge Proof ( + HomEIG‐ZKProof) and the national encryption algorithm SM2 is proposed. While ensuring transaction traceability and verifiability, this scheme achieves privacy protection for both payers’ and recipients’ identities, specifically tailored for cross‐border trade scenarios. Additionally, customs authorities play the role of supervisory nodes to verify the identities of transaction parties and the zero‐knowledge proofs for transaction information. The RAFT consensus algorithm is employed to construct a secure authentication application, demonstrating how zero‐knowledge proofs, combined with homomorphic encryption, can be verified through a consensus process. In this scenario, the legitimacy of transaction amounts is subject to zero‐knowledge verification during consensus interactions. Merchant identity verification is accomplished using SM2 ring signatures. The analysis indicates that this scheme offers strong security features such as resistance to tampering attacks, public key replacement attacks, impersonation attacks, and anonymity. Testing results demonstrate that this scheme can effectively provide dual privacy protection for transaction amounts and identities in cross‐border trade, meeting the practical requirements of privacy protection in cross‐border trade transactions.
Andrew Jeffery, Julien Maffre, Heidi Howard, Richard Mortier
Software services are increasingly migrating to the cloud, requiring trust in actors with direct access to the hardware, software and data comprising the service. A distributed datastore storing critical data sits at the core of many services; a prime example being etcd in Kubernetes. Trusted execution environments can secure this data from cloud providers during execution, but it is complex to build trustworthy data storage systems using such mechanisms. We present the design and evaluation of the Ledger-backed Secure Key-Value datastore (LSKV), a distributed datastore that provides an etcd-like API but can use trusted execution mechanisms to keep cloud providers outside the trust boundary. LSKV provides a path to transition traditional systems towards confidential execution, provides competitive performance compared to etcd, and helps clients to gain trust in intermediary services. LSKV forms a foundational core, lowering the barriers to building more trustworthy systems.
In this era of rapid technological evolution and ever-changing regulatory landscapes, this chapter explores the transformative journey of data accountability. Beginning with an exploration of data accountability in cloud environments, the narrative sets the stage by defining its scope and surveying the current research landscape. Navigating the intricate terrain of regulatory compliance in the cloud, the discussion unfolds with a focus on prominent regulations such as General Data Protection Regulation (GDPR) and California Consumer Privacy Act (CCPA). A comparative analysis sheds light on the varying compliance requirements and the consequential impact on Cloud Service Providers. This chapter then navigates through diverse accountability models in cloud computing, reevaluating the shared responsibility model and introducing advanced frameworks. It advocates for the integration of accountability directly into the fabric of cloud service level agreements. An in-depth examination of advanced auditing and monitoring mechanisms follows suit, featuring real-time auditing tools, log analysis techniques, and insights into monitoring tool kits offered by major cloud platforms like Google Cloud and AWS. Grounding the discourse in practicality, the narrative showcases case studies that illuminate successful implementations of data accountability. These narratives not only celebrate achievements but also outline encountered challenges and the strategies employed for effective mitigation. A comparative analysis of accountability practices in the realms of Google Cloud and AWS enriches the practical insights. Looking ahead, the narrative concludes with a forward-thinking exploration of future trends in data accountability. It envisions the application of machine learning for anomaly detection, the integration of blockchain and distributed ledger technologies, and anticipates the evolving landscape of regulatory frameworks. As a compass for professionals, researchers, and policymakers, this text serves as an invaluable guide through the dynamic intersection of technology and data governance.
Frederico Baptista, Marina Dehez-Clementi, Jonathan Detchart
The integration of Unmanned Aircraft Systems (UASs) into the current airspace poses significant challenges in terms of safety, security, and operability. As an example, in 2019, the European Union defined a set of rules to support the digitalization of UAS traffic management (UTM) systems and services, namely the U-Space regulations. Current propositions opted for a centralized and private model, concentrated around governmental authorities (e.g., AlphaTango provides the Registration service and depends on the French government). In this paper, we advocate in favor of a more decentralized and transparent model in order to improve safety, security, operability among UTM stakeholders, and legal compliance. As such, we propose DFly, a publicly auditable and privacy-preserving UAS traffic management system on Blockchain, with two initial services: Registration and Flight Authorization. We demonstrate that the use of a blockchain guarantees the public auditability of the two services and corresponding service providers’ actions. In addition, it facilitates the comprehensive and distributed monitoring of airspace occupation and the integration of additional functionalities (e.g., the creation of a live UAS tracker). The combination with zero-knowledge proofs enables the deployment of an automated, distributed, transparent, and privacy-preserving Flight Authorization service, performed on-chain thanks to the blockchain logic. In addition to its construction, this paper details the instantiation of the proposed UTM system with the Ethereum Sepolia’s testnet and the Groth16 ZK-SNARK protocol. On-chain (gas cost) and off-chain (execution time) performance analyses confirm that the proposed solution is a viable and efficient alternative in the spirit of digitalization and offers additional security guarantees.
Jialing He, Ning Wang, Tao Xiang, Yiqiao Wei · 7 authors
While smart grid significantly facilitates energy efficiency by using users’ power consumption data, it poses privacy leakage risk for user personal behaviors. Differential privacy (DP) has emerged as a promising solution to address this issue. However, existing approaches suffer from severe data utility degradation due to the intensive noise introduced by DP. Additionally, some of these methods are vulnerable to security attacks. To bridge this gap, in this paper, we propose ABDP (accuratebilling-enableddifferentiallyprivate), a mechanism that achieves high-strength DP while ensuring accurate aggregation and billing operations without compromising security. In particular, we propose aggregated and individual noise cancellation algorithms to counteract the negative effects of noise on data utility. Specifically, our ABDP ensures precise aggregation and accurate billing calculations for the power grid and individual users, respectively Furthermore, we present a Blockchain smart contract exploiting the pseudo random function to enforce a fair and secure data reporting process. Theoretical analysis is provided to evaluate the privacy and security guarantees of ABDP. Experimental results on real-world datasets, namely NERL-DATA and REDD, demonstrate that ABDP achieves error-free aggregation and billing calculation, offers arbitrary intensity privacy protection against non-intrusive load monitoring and filtering attacks, and outperforms existing state-of-the-art approaches.
This paper introduces a novel framework for verifiable privacy-preserving computation on the blockchain, addressing the challenges of managing confidential transactions within the transparent and distributed ledger. The proposed framework leverages Secure Multi-Party Computation (MPC) and integrates it with blockchain technology to ensure both privacy and public verifiability of computation results. Notably, in our framework, the primary computation occurs off-chain to avoid the inefficiencies of smart contracts, while the results are uploaded to the blockchain for immutability and public verification. Furthermore, through the adoption of a zkRollup-like design and the selection of an efficient MPC algorithm, our framework demonstrates robust structural scalability and low communication complexity compared to other similar approaches.
Sizai Hou, Songze Li, Tayyebeh Jahani-Nezhad, Giuseppe Caire
Federated learning (FL) has recently gained significant momentum due to its potential to leverage large-scale distributed user data while preserving user privacy. However, the typical paradigm of FL faces challenges of both privacy and robustness: the transmitted model updates can potentially leak sensitive user information, and the lack of central control of the local training process leaves the global model susceptible to malicious manipulations on model updates. Current solutions attempting to address both problems under the one-server FL setting fall short in the following aspects: 1) designed for simple validity checks that are insufficient against advanced attacks (e.g., checking norm of individual update); and 2) partial privacy leakage for more complicated robust aggregation algorithms (e.g., distances between model updates are leaked for multi-Krum). In this work, we formalize a novel security notion of aggregated privacy that characterizes the minimum amount of user information, in the form of some aggregated statistics of users' updates, that is necessary to be revealed to accomplish more advanced robust aggregation. We develop a general framework PriRoAgg, utilizing Lagrange coded computing and distributed zero-knowledge proof, to execute a wide range of robust aggregation algorithms while satisfying aggregated privacy. As concrete instantiations of PriRoAgg, we construct two secure and robust protocols based on state-of-the-art robust algorithms, for which we provide full theoretical analyses on security and complexity. Extensive experiments are conducted for these protocols, demonstrating their robustness against various model integrity attacks, and their efficiency advantages over baselines.
Federated learning (FL) enables distributed clients to privately train machine learning models using their own local data, thus avoiding the security risk of directly exchanging private data between clients and servers. However, there is a risk in federated learning: a malicious party may launch a reverse attack based on the parameters uploaded by the client to analyze the attributes of the client’s local private data. Some studies use the Shamir secret sharing method and zero-knowledge proofs (ZKP) to ensure the privacy and integrity of the inputs in FL. However, the Shamir secret sharing scheme often requires an optimistic guarantee on the number of malicious clients, which cannot obtain a sufficient number of secret slices through collusion, and the general ZKP scheme requires the clients to compute the proof belongs to others during the proof and verification stages, and its efficiency needs to be improved. In this paper, we propose VSSPFL method to achieve efficient and secure data collaboration, which improves the Shamir secret sharing scheme to prevent reverse attacks by an unknown number of malicious clients and reduces the cost of ZKP by using a probabilistic integrity check method.