Smart cities increasingly rely on the Internet of Things (IoT) to enhance infrastructure and public services. However, many existing IoT frameworks face challenges related to security, privacy, scalability, efficiency, and low latency. This paper introduces the Blockchain and Federated Learning for IoT (BFLIoT) framework as a solution to these issues. In the proposed method, the framework first collects real-time data, such as traffic flow and environmental conditions, then normalizes, encrypts, and securely stores it on a blockchain to ensure tamper-proof data management. In the second phase, the Data Authorization Center (DAC) uses advanced cryptographic techniques to manage secure data access and control through key generation. Additionally, edge computing devices process data locally, reducing the load on central servers, while federated learning enables distributed model training, ensuring data privacy. This approach provides a scalable, secure, efficient, and low-latency solution for IoT applications in smart cities. A comprehensive security proof demonstrates BFLIoT’s resilience against advanced cyber threats, while performance simulations validate its effectiveness, showing significant improvements in throughput, reliability, energy efficiency, and reduced delay for smart city applications.
Attribute Based Access Control (ABAC) is one the most efficient, scalable, and well used access control. It’s based on attributes not on users, but even when the users want to get access to some resource, they must submit their attributes for the verification process which may reveal the privacy of the users. Many research papers suggest blockchain-based ABAC which provides an immutable and transparent access control system. However, the privacy of the system may be compromised depending on the nature of the attributes. A Zero-Knowledge Proof, Ethereum-Based Access Control (ZK‑ABAC) is proposed in this paper to simplify the management of access to the devices/objects and provide an efficient and immutable platform that keeps track of all actions and access management and preserve the privacy of the attributes. Our ZK-ABAC model utilizes smart contracts to facilitate access control management, Zero-Knowledge Succinct NonInteractive Argument of Knowledge (ZK-SNARK) protocol to add privacy to attributes, InterPlanetary File System (IPFS) network to provide distributed storage system, and Chainlink to manage communications and data between on/ off-chain systems. Comprehensive experiments and tests were conducted to evaluate the performance of our model, including the implementation of ZK-SNARK on the Ethereum blockchain. The results demonstrated the scalability challenges in the setup and proving phases, as well as the efficiency gains in the verification phase, particularly when scaled to higher numbers of users. These findings underscore the practical viability of our ZK-ABAC model for secure and privacy-preserving access control in decentralized environments.
Timed Data Release (TDR) is a practical security mechanism that safeguards data until a prescribed time has elapsed. However, existing TDR frameworks do not focus on reliability guarantees and lack formal security analysis. To this end, we propose PR-TDR, a novel framework that supports privacy-preserving and reliable timed data release while providing provable security properties. PR-TDR includes two novel contributions: a formal privacy-preserving design for TDR, named P-TDR and a reliable lifetime secret key management built on top of P-TDR that systematically empowers P-TDR with reliability. P-TDR prevents adversaries from accessing the data prior to the prescribed release time. At the core of the design of P-TDR, a group of decentralized peers, which operates under an honest-majority assumption, collaboratively takes charge of managing the lifetime secret key. Each peer stores a key share of the secret key. The proposed reliability layer that empowers P-TDR with reliability guarantees incorporates two carefully designed protocols that operate before the prescribed release time, namely the lifetime secret key auditing protocol and the lifetime secret key share recovery protocol. The auditing protocol enables a semi-honest auditor to confirm the availability of the lifetime secret key with the peers while not gaining any knowledge about the secret key itself. The recovery protocol allows peers that have lost their respective shares of the lifetime secret key to recover them with the help of other peers, ensuring that the lifetime secret key remains private. We provide formal security proof to demonstrate that PR-TDR satisfies the desired security properties. We implement our framework using Ethereum and our performance evaluations confirm that PR-TDR not only embodies the desired security properties but also operates efficiently.
Blockchain technology has significantly transformed the landscape of financial transactions, offering unparalleled transparency and trust. However, the need for privacy in recurring transactions remains a critical challenge. This paper proposes an innovative approach to enhancing privacy in blockchain-based recurring transactions through the integration of Zero-Knowledge Proofs (ZKPs). We present a detailed architecture that combines the efficiency of smart contracts on the Ethereum blockchain with the confidentiality offered by ZKPs. Our implementation ensures that transaction details remain private while maintaining the integrity and non-repudiation inherent to blockchain technology. We evaluate the proposed solution in terms of performance, scalability, and security, demonstrating its potential to revolutionize privacy-preserving recurring payments in decentralized finance (DeFi) applications. This work addresses the privacy paradox in blockchain and sets the stage for future advancements in secure, automated financial systems.
Lin Chen, Yuxiang Chen, Wei Liang, Xiong Li · 7 authors
With the swift advancement of the Internet of Things (IoT) and artificial intelligence (AI), various technologies have been integrated into wearable medical health devices, improving users’ awareness of their physical states and enabling the analysis of a greater amount of human data. However, these sensitive pieces of information are prone to tampering or theft during storage and transmission, posing security risks. In this article, we propose a multiattribute sketch secure data sharing scheme for IoT wearable medical devices based on blockchain (MASS). We introduce a multiattribute sketch storage method that stores the encrypted hash of health data transmitted by medical wearable devices on the blockchain. This work also designs a ciphertext-policy attribute-based encryption (CP-ABE) access control mechanism that effectively addresses the secure sharing of data from wearable medical devices among healthcare professionals. Experimental findings indicate that with the rise in the number of medical health data documents, the costs associated with index generation and search time decrease by 55.3% and 10.83%, respectively. Additionally, as the frequency of data access increases, there is a 13.5% reduction in encryption time, and the implementation of multiattribute sketches results in a 24.8% and 11.3% reduction in index generation and search times, respectively.
Introduced by Google in 2016, federated learning (FL) is a distributed machine learning framework to ensure data privacy amid the surge in big data. FL enables secure data sharing without accessing local data. Despite its advantages, it faces challenges due to the limited participation of the data owner. To address this, this article proposes the model value transfer incentive (MVTI) to enhance FL incentives for Artificial Intelligence of Things (AIoT). MVTI allows active participation of data requesters in FL training, addressing limited data owner engagement, and facilitating personalized model construction. The integrated model bail and contribution assessment mechanism ensures fair benefit redistribution. Using smart contracts (SCs) and interplanetary file system (IPFS) enhances security and reliability, ensuring transparent and tamper-resistant execution for secure transactions and data integrity. Our experiments highlight MVTI’s superiority in addressing FL incentive challenges for AIoT compared to state-of-the-art baselines on real-world datasets. We also demonstrate the compatibility of multiple gradient protections with incentive mechanisms, especially with gradient compression. The proposed SC-MVTI scheme is resilient and demonstrates the potential to significantly improve the overall efficacy of the FL system within incentive frameworks.
This article proposes a novel method for managing usage counters within an anonymous credential system, addressing the limitation of traditional anonymous credentials in tracking repeated use. The method takes advantage of blockchain technology through Smart Contracts deployed on the Ethereum network to enforce a predetermined maximum number of uses for a given credential. Users retain control over increments by providing zero-knowledge proofs (ZKPs) demonstrating private key possession and agreement on the increment value. This approach prevents replay attacks and ensures transparency and security. A prototype implementation on a private Ethereum blockchain demonstrates the feasibility and efficiency of the proposed method, paving the way for its potential deployment in real-world applications requiring both anonymity and usage tracking.
In modern healthcare, providers increasingly use cloud services to store and share electronic medical records. However, traditional cloud hosting, which depends on intermediaries, poses risks to privacy and security, including inadequate control over access, data auditing, and tracking data origins. Additionally, current schemes face significant limitations such as scalability concerns, high computational overhead, practical implementation challenges, and issues with interoperability and data standardization. Unauthorized data access by cloud providers further exacerbates these concerns. Blockchain technology, known for its secure and decentralized nature, offers a solution by enabling secure data auditing in sharing systems. This research integrates blockchain into healthcare for efficient record management. We proposed a blockchain-based method for secure EHR management and integrated Ciphertext-Policy Attribute-Based Encryption (CP-ABE) for fine-grained access control. The proposed algorithm combines blockchain and smart contracts with a cloud-based healthcare Service Management System (SMS) to ensure secure and accessible EHRs. Smart contracts automate key management, encryption, and decryption processes, enhancing data security and integrity. The blockchain ledger authenticates data transactions, while the cloud provides scalability. The SMS manages access requests, enhancing resource allocation and response times. A dual authentication system confirms patient keys before granting data access, with failed attempts leading to access revocation and incident logging. Our analyses show that this algorithm significantly improves the security and efficiency of health data exchanges. By combining blockchain's decentralized structure with the cloud's scalability, this approach significantly improves EHR security protocols in modern healthcare setting.
DNA fingerprinting and matching for identifying suspects has been a common practice in criminal investigation. Such proceedings involve multiple parties such as investigating agencies, suspects and forensic labs. A major challenge in such settings is to carry out the matching process between the suspects' DNA samples and the samples obtained from the crime scene without compromising the privacy of the suspects' DNA profiles. Additionally, it is necessary that sensitive details pertaining to the investigation such as the identities of the suspects and evidence obtained from the crime scene must be kept private to the investigating agency. We present a novel DNA matching scheme, termed as PrivaMatch, which addresses multiple concerns about privacy of the suspects' DNA profiles and the crime scene evidence. In the proposed scheme, the investigating agencies oblivious transfer and zero-knowledge proofs to privately obtain the DNA profiles of the suspects from the forensic lab's database.In addition, we present a clever data obfuscation technique using homomorphic encryption and modular arithmetic for the investigating agency to privately obtain the DNA profile of the crime scene's sample, keeping the profile oblivious from the forensic lab. The DNA profile of the crime scene sample is operated on using a homomorphic cryptosystem such that neither of the parties (e.g., the investigation agency, forensic labs, DNA database owners) learns about the private data of the other parties. The proposed scheme is analysed formally and the practicality of its security strengths is verified using simulations under standard assumptions.
Nicolas Huber, Ralf Küsters, Julian Liedtke, Daniel Rausch
Abstract Electronic voting (e-voting) systems have become more prevalent in recent years, but security concerns have also increased, especially regarding the privacy and verifiability of votes. As an essential ingredient for constructing secure e-voting systems, designers often employ zero-knowledge proofs (ZKPs), allowing voters to prove their votes are valid without revealing them. Invalid votes can then be discarded to protect verifiability without compromising the privacy of valid votes. General purpose zero-knowledge proofs (GPZKPs) such as ZK-SNARKs can be used to prove arbitrary statements, including ballot validity. While a specialized ZKP that is constructed only for a specific election type/voting method, ballot format, and encryption/commitment scheme can be more efficient than a GPZKP, the flexibility offered by GPZKPs would allow for quickly constructing e-voting systems for new voting methods and new ballot formats. So far, however, the viability of GPZKPs for showing ballot validity for various ballot formats, in particular, whether and in how far they are practical for voters to compute, has only recently been investigated for ballots that are computed as Pedersen vector commitments in an ACM CCS 2022 paper by Huber et al. Here, we continue this line of research by performing a feasibility study of GPZKPs for the more common case of ballots encrypted via Exponential ElGamal encryption. Specifically, building on the work by Huber et al., we describe how the Groth16 ZK-SNARK can be instantiated to show ballot validity for arbitrary election types and ballot formats encrypted via Exponential ElGamal. As our main contribution, we implement, benchmark, and compare several such instances for a wide range of voting methods and ballot formats. Our benchmarks not only establish a basis for protocol designers to make an educated choice for or against such a GPZKP, but also show that GPZKPs are actually viable for showing ballot validity in voting systems using Exponential ElGamal.
Abeera Malik, Talha, Muhammad Zunnurain Hussain, Muzammil Mustafa · 7 authors
The Internet of Things (IoT) has become a foundational element of the digital infrastructure, extending its connectivity across various sectors and embedding intelligence in everyday devices. This article introduces SecureNet, a pioneering approach that integrates Machine Learning (ML), Blockchain, and Federated Learning (FL) to enhance IoT security. To navigate this challenging train, an innovative framework that synergizes Machine Learning (ML), Blockchain technology, and Federated Learning (FL) to fortify IoT security. SecureNet is architected to deliver a robust defense mechanism for IoT ecosystems, providing resilience against increasingly sophisticated cyber threats, and ensuring the preservation of data integrity, privacy, and unwavering system reliability. This study explores the application of advanced ML techniques NSL-KDD dataset, implementing two highly effective classifiers: Random Forest and Logistic Regression. The Random Forest classifier exhibited an exceptional accuracy of 99.85%, while the Logistic Regression model demonstrated a near-perfect accuracy of 99.03%. These compelling results highlight the efficacy of ML in identifying and mitigating activities within network traffic. SecureNet leverages ML’s profound analytical capabilities for intelligent threat discernment, Blockchain’s immutable ledgers for unassailable data verification, and FL’s privacy-centric approach to distribute model training. These outcomes underscore the potential of ML models to enhance IoT security by accurately identifying malicious patterns and anomalies within network traffic.
With the exponential growth of smart gadgets and the technologies that connect them, the Internet of Things (IoT) has emerged as the most promising new technology of the last decade. This is true from both a business and academic standpoint. Deployment in smart-home and smart-city initiatives throughout the globe has increased IoT's popularity. Regrettably, the processing power, storage capacity, and bandwidth of IoT network devices are severely constrained. So, compared to other endpoint devices like PCs, tablets, and smartphones, they are more susceptible to assaults. The development of a safe and reliable model for commonly used operations is booming, including data collection, processing, storage, and communication. It isn't easy to build such a model since there are so many obstacles to overcome. New cybersecurity risks emerge on a regular basis, while old ones remain dormant and ready to be exploited. Sadly, these cyber dangers also serve to safeguard the IoT. Due to recent developments, research into cyberphysical systems, the internet of things (IoT), and blockchain is more important than ever. In Internet of Things (IoT) systems, where security breaches compromise the centralised system's ability to function, blockchain technology has shown remarkable promise. Distributed ledgers powered by blockchain technology may enhance the security of the Internet of Things (IoT) and its user access management systems, making them more reliable. Following the safe vehicle authorization, the suggested model presents the idea of a branching blockchain that considers the most recently utilized block. This allows low-powered devices to also benefit from the blockchain's security. The needs of the IoV at the physical layer are beyond the capabilities of present-day mining technology. The suggested architecture has many important aspects, such as a decentralized ledger system, load balancing, scalability, and a lightweight blockchain for the physical layer and above. The comparison investigation, which takes computation and transmission costs into account, shows that our proposed framework performs better than the ones that are presently in use. If anybody is interested in a lightweight blockchain variation that can be applied to the physical layer of the IoV, this model may serve as the foundation for new research areas.
The security of academic credentials is increasingly at risk due to cyberattacks and credential fraud. Traditional verification systems rely on centralised databases, creating single points of failure and privacy concerns. This paper explores Zero-Knowledge Proofs (ZKPs) with blockchain to enhance educational data security. We propose an optimised ZKP protocol for education, improving lightweight infrastructure, scalability, credential revocation, and ease of use. By refining existing techniques, our approach enhances secure, privacy-preserving credential verification, creating a resilient educational data system.
Federated learning, as a distributed machine learning method is facing significant challenges in data privacy protection at present. This study aims to enhance the privacy protection capability of federated learning by integrating blockchain consensus mechanisms, as well as proposing a novel solution by designing a decentralized federated learning framework using consensus mechanisms such as Proof of Stake (PoS), and implementing smart contracts to automate key processes. The research results demonstrate that this approach effectively enhances the security and transparency of data processing while ensuring the efficiency and scalability of the system. Furthermore, the adoption of homomorphic encryption technology further ensures the security and integrity of data during transmission.
Abstract: In traditional digital identity management solutions, the authentication process mostly adopts a centralized authentication model, where user identity data is stored in a single authentication authority, which leads to a myriad of problems such as data leakage, data tampering, and single point of failure. How to ensure the security of user identity data and how to carry out effective privacy protection is the key to ensuring the digitalization process in various industries. Blockchain has the characteristics of decentralization and tampering, which can provide a reliable data storage solution. Zero-knowledge proof technology can verify the correctness of certain facts without revealing specific information. In digital identity applications, zero-knowledge proof allows users to prove their identity without revealing any personal information. Therefore, this paper proposes an identity authentication model based on blockchain and zero-knowledge proof to ensure the security of the user's identity data during the identity authentication process. Meanwhile, in order to make the authentication process more reliable, smart contract technology is introduced to complete the whole authentication process.
Olive Franzese, Ali Shahin Shamsabadi, Luck, Carter, Haddadi, Hamed
The black-box service model enables ML service providers to serve clients while keeping their intellectual property and client data confidential. Confidentiality is critical for delivering ML services legally and responsibly, but makes it difficult for outside parties to verify important model properties such as fairness. Existing methods that assess model fairness confidentially lack either (i) reliability because they certify fairness with respect to a static set of data, and therefore fail to guarantee fairness in the presence of distribution shift or service provider malfeasance; and/or (ii) scalability due to the computational overhead of confidentiality-preserving cryptographic primitives. We address these problems by introducing online fairness certificates, which verify that a model is fair with respect to data received by the service provider online during deployment. We then present OATH, a deployably efficient and scalable zero-knowledge proof protocol for confidential online group fairness certification. OATH exploits statistical properties of group fairness via a cut-and-choose style protocol, enabling scalability improvements over baselines.
Artificial intelligence (AI) technologies have been applied to the Internet of Vehicles (IoV) to provide convenience services such as traffic flow prediction. However, concerns regarding privacy and security are on the rise as huge amounts of data are aggregated to form large models (LMs). Although federated learning (FL), which trains and updates a model without sharing the actual datasets, has been intensively researched to prevent privacy breaches, there are still potential security threats like a single point of failure and intentional tampering with malicious data. This is because of the vulnerability of a central curator and a lack of authentication. As participants, they (i.e., vehicles) may unintentionally update low-quality data caused by poor wireless connectivity, unstable availability, and insufficient training datasets. They may also intentionally update unreliable data to carry out poisoning attacks. The divergence among local models, trained on non-independent and identically distributed (non-IID) data, can slow convergence and diminish model accuracy when these models are aggregated. Therefore, it is important to carefully select trustworthy participants. In this paper, we propose a new reliable and secure federated learning for IoV based on decentralized blockchain and reputation management. To cope with a single point of failure, injection of malicious data, and lack of authentication while ensuring privacy and traceability, our scheme combines blockchain and a lightweight digital signature. Moreover, we employ the concept of the reputation of vehicles to select suitable participants with reliability, ultimately improving accuracy. Security analysis results, including comparisons with previous works, prove that the proposed scheme can address security concerns. The results of performance evaluations demonstrate the effectiveness of our proposed scheme.
Alberto Ballesteros-Rodríguez, Salvador Sánchez‐Alonso, Miguel‐Ángel Sicilia
Abstract The growing adoption of wearables is driving the demand for personalized services that leverage unprocessed data, such as biometric and health information, to enhance user experiences and support through software applications. However, several existing use cases involving this information still prioritize traditional schemes, neglecting user privacy. Consequently, the transparency of data transmission paths and the potential for tampering remain ambiguous when users share data with service providers. In this paper, we propose the application of an Internet of Things device-focused distributed ledger as an underlying layer for the transmission of encrypted data using streams. Moreover, our proposal enables data recording for future events and the implementation of multi-subscriber models, allowing client information to be shared securely with different service providers. Through simulation experiments conducted on constrained devices, we demonstrate that our proposed framework efficiently transmits large ciphertexts through streams on a distributed ledger, overcoming the inherent limitations of such networks when dealing with substantial data volumes. Ultimately, the performance metrics presented prove that the proposed model is suitable for real-world applications requiring continuous data collection by wearables and subsequent transmission to service providers.
R. Shashidhara, Renju Chirakarotu Nair, Pavan Kumar Panakalapati
ABSTRACT Blockchain technology has revolutionized data sharing and storage, yet maintaining privacy and security while ensuring transparency and immutability remains a significant challenge. This research addresses this challenge by investigating the promising solution of Zero‐Knowledge Proofs (ZKPs), which allow for validation without revealing additional information. Unlike previous studies that focus on specific ZKP types or limited performance metrics, this article provides a comprehensive analysis of multiple ZKP technologies, along with their distinct features and blockchain applications. We evaluate ZKP tools such as snarkjs, ZoKrates, and Circom, focusing on key factors including proof size, trusted setup, prover time, verifier time, and scalability. A detailed case study on the Ethereum blockchain demonstrates the practical implementation and performance of ZKPs, effectively bridging the gap between theoretical analysis and real‐world applications. Furthermore, we explore how ZKPs can significantly enhance scalability in blockchain applications. By proposing innovative ZKP use cases and addressing current challenges, this article not only provides a comprehensive understanding of ZKP technologies but also outlines potential future research directions, making a significant impact on the field of blockchain privacy and security.
The increased use of artificial intelligence generated content (AIGC) among vast user populations has heightened the risk of private data leaks. Effective auditing and regulation remain challenging, further compounding the risks associated with the leaks involving model parameters and user data. Blockchain technology, renowned for its decentralized consensus mechanism and tamper-resistant properties, is emerging as an ideal tool for documenting, auditing, and analyzing the behaviors of all stakeholders in machine learning as a service (MLaaS). This study centers on biometric recognition systems, addressing pressing privacy and security concerns through innovative endeavors. We conducted experiments to analyze six distinct deep neural networks, leveraging a dataset quality metric grounded in the query output space to quantify the value of the transfer datasets. This analysis revealed the impact of imbalanced datasets on training accuracy, thereby bolstering the system's capacity to detect model data thefts. Furthermore, we designed and implemented a novel Bio-Rollup scheme, seamlessly integrating technologies such as certificate authority, blockchain layer two scaling, and zero-knowledge proofs. This innovative scheme facilitates lightweight auditing through Merkle proofs, enhancing efficiency while minimizing blockchain storage requirements. Compared to the baseline approach, Bio-Rollup restores the integrity of the biometric system and simplifies deployment procedures. It effectively prevents unauthorized use through certificate authorization and zero-knowledge proofs, thus safeguarding user privacy and offering a passive defense against model stealing attacks.
Xuanxia Yao, Jinyuan Zhou, Xiaojiang Du, Shurong Zhang
Nowadays, we are living in an open network environment with varieties of smart devices, which makes individual privacy face unprecedented threats. For one thing, a plenty of sensitive information may be gathered without the owner’s knowledge. For the other, the Internet of Things (IoT)-based services and various intelligent applications require a large amount of perceptual data. And in practice, these data are usually encrypted and stored in storage providers like cloud for security and cost saving. To fully harness the productivity value of data and protect privacy, ciphertext-policy attribute-based encryption (CP-ABE) is widely used. Nevertheless, most existing CP-ABE schemes cannot work well for IoT because of the heavy overhead and the open and distributed environment. To lower the cost, a lightweight CP-ABE scheme without pairing is proposed and proved in the set-selective mode. Both the theoretical analysis and experiments show its advantages in computation, communication, and storage overhead. For flexible access control in IoT, we attempt to employ the masked authenticated message (MAM) mechanism of the IOTA to manage authorization for our CP-ABE scheme. Comparisons with similar schemes show that it can overcome the low throughput and monetary cost in other distributed ledger-based access control schemes.
This paper proposes a new global model aggregation method based on using zero-knowledge federated learning (ZKFL). The purpose is to secure horizontal or P2P federated machine learning systems with shorter aggregation times, higher model accuracy, and lower system costs. We use a model parameter-sharing Chord overlay network among all client hosts. The overlay guarantees a trusted sharing of zero-knowledge proofs for aggregation integrity, even under malicious Byzantine attacks. We tested over popular datasets, Fashion-MNIST and CIFAR10, to prove the new system protection concept. Our benchmark experiments validate the claimed advantages of the ZKFL scheme in all objective functions. Our aggregation method can be applied to secure both rank-based and similarity-based aggregation schemes. For a large system with over 200 clients, our system takes only 3 seconds to yield high-precision global machine models under the ALIE attacks with the Fashion-MNIST dataset. We have achieved up to 85% model accuracy, compared to only 3%$\sim$45% accuracy observed with federated schemes without protection. Moreover, our method demands a low memory overhead for handling zero-knowledge proofs as the system scales greatly to a larger number of client nodes.
Alexandre Pacheco, Sébastien De Vos, Andreagiovanni Reina, Marco Dorigo · 5 authors
Federated learning is a new approach to distributed machine learning that offers potential advantages such as reducing communication requirements and distributing the costs of training algorithms. Therefore, it could hold great promise in swarm robotics applications. However, federated learning usually requires a centralized server for the aggregation of the models. In this paper, we present a proof-of-concept implementation of federated learning in a robot swarm that does not compromise decentralization. To do so, we use blockchain technology to enable our robot swarm to securely synchronize a shared model that is the aggregation of the individual models without relying on a central server. We then show that introducing a single malfunctioning robot can, however, heavily disrupt the training process. To prevent such situations, we devise protection mechanisms that are implemented through secure and tamper-proof blockchain smart contracts. Our experiments are conducted in ARGoS, a physics-based simulator for swarm robotics, using the Ethereum blockchain protocol which is executed by each simulated robot.
With the help of artificial intelligence, the large amount of data generated by Internet of Things (IoT) has unleashed significant value. Federated learning is emerging as a novel paradigm which can be applied to solve the privacy issues caused by analyzing IoT data. However, traditional federated learning protocols are vulnerable to inference and poisoning attacks. Various solutions have been proposed to enhance data privacy and robustness. Nonetheless, most of these solutions are usually centralized and rely on unrealistic security assumptions. Furthermore, the recently proposed blockchain-based decentralized solutions generally incur high costs, which is unaffordable for resource-constrained IoT devices. In this article, we propose a practical secure federated learning system named PrSeFL. We utilize blockchain to decentralize the federated learning process so that the security assumptions are easier to achieve in practice. To preserve data privacy, we implement secure multiparty computation-based secure aggregation in blockchain environment. To guarantee practical robustness, we enforce norm constraints on the masked updates via zero-knowledge proof. Moreover, we propose a modified dynamic accumulator which is utilized to realize lightweight anonymous authentication of users. Simulation results show that, compared with state-of-the-art systems, PrSeFL has superior performance on authentication and model training. And the advantage of PrSeFL becomes more significant as the number of users grows.