Byzantine Fault Research Articles

Asynchronous federated learning with directed acyclic graph-based blockchain in edge computing: Overview, design, and challenges

Asynchronous Federated Learning (AFL) has been introduced to improve the efficiency of FL by reducing the latency of Machine Learning (ML) model aggregation, particularly in the Internet of Things (IoT) environment. Meanwhile, decentralized FL, such as leveraging blockchain and directed acyclic graph (DAG)-based ledgers, also has drawn much attention to the integration of FL owing to the security benefit against single-point-failure and Byzantine fault tolerant consensus. We observe that the inherent network asynchrony of DAG-based ledgers is beneficial for implementing asynchronous FL particularly in edge computing domains, even though there is no existing survey work that provides a fundamental overview of such integration. This paper surveys on the integration of asynchronous FL with DAG, which we call AFL-DAG, as a promising approach to realize the intersection of decentralized FL and asynchronous FL. Motivated by the lack of a concrete taxonomy of asynchronous FL and a global model, especially with decentralized FL, we introduce universally applicable terminologies and the extensive classification method of FL in terms of asynchrony. Based on the proposed taxonomy, we present a generic system model of AFL-DAG for edge computing applications. To provide a horizontal overview and cover fundamental concepts in AFL-DAG, we identify four critical design factors and the state-of-the-art solutions are discussed accordingly. Future directions to achieve a practical AFL-DAG are also highlighted. Finally, we explore the opportunities of AFL-DAG by investigating a popular edge computing application, the on-device crowdsourcing, and provide a high-level evaluation of AFL-DAG compared to blockchained-FL.

Reinforced practical Byzantine fault tolerance consensus protocol for cyber physical systems

Cyber–physical systems (CPS) are becoming an essential component of modern life. CPS services enable information to be exchanged between physical devices and virtual systems. The increasing malicious activities causing confidential data leakage and incorrect performance of devices are posing challenges for protection against cyber threats. Therefore, development of effective solutions that can protect both CPS data and data exchange networks are extremely urgent. This study makes some improvements on Practical Byzantine Fault Tolerance (PBFT), and provides a critical analysis of the feasibility of using blockchain technology to protect constrained CPS data. It also justifies the choice of the improved PBFT for implementation on such devices and simulates the main distributed ledger scenarios. The improved PBFT works as follows: first, the client broadcasts the signed transaction data to the entire network, rather than just the master node, followed by a hash value comparison verification process; second, select the master node based on the longest chain principle, and punish the malicious node via the “blacklist” mechanism; third, add the data synchronization and verification process, as well as dynamically entering and leaving nodes via state transitions. Ultimately, the PBFT commit phase is eliminated, resulting in a two-stage process. In comparison to the original PBFT protocol, the Reinforced Practical Byzantine Fault Tolerance (RPBFT) protocol may substantially enhance system throughput and minimize consensus communication time, thereby improving overall system efficiency while guaranteeing security.

Secure PBFT Consensus-Based Lightweight Blockchain for Healthcare Application

Recent advancement in IoT technology has boosted the healthcare domain with enormous usage of IoT devices to provide elevated services to patients with chronic disorders on a real-time basis by the incorporation of IoT sensors on patients’ bodies. However, providing services ensuring security and maintaining the privacy of patients is a challenging task. Blockchain technology promises security in a distributed environment but popular consensus algorithms such as Proof of Work (PoW) and Proof of Stake (PoS) require huge computational resources and energy by making the IoT environment inefficient. This paper introduces a secure Practical Byzantine Fault Tolerance (PBFT) consensus-based lightweight blockchain algorithm for healthcare applications. To strengthen the PBFT consensus, highly trusted nodes were allowed to participate in the consensus algorithm using the Eigen Trust model and Verifiable Random Function (VRF) to select a random primary node from a group of trusted consensus nodes. The proposed algorithm is tested in a simulated environment and evaluated against the traditional PBFT consensus algorithm considering throughput, latency, and fault tolerance.

A Parallel Consensus Mechanism Using PBFT Based on DAG-Lattice Structure in the Internet of Vehicles

The blockchain technology in the Internet of Vehicles (IoV) can be applied to solve security problems. Due to the high dynamics and resource constraints of the IoV, the traditional single-chain blockchain system is not suitable for the practical scenarios. In our previous work, we proposed a lightweight blockchain architecture based on the directed acyclic graph lattice (DAG-lattice) structure for the IoV. However, the current consensus algorithms are not suitable for the previous architecture mentioned above. Therefore, this article proposes a parallel consensus mechanism based on the DAG-lattice structure in the IoV using practical Byzantine fault tolerance (PBFT) as consensus algorithm. This mechanism solves the problem of low efficiency of PBFT caused by too many consensus nodes and the mobility of nodes by adopting the method of network sharding, and meanwhile to ensure that transactions can be reached on consensus parallelly. To avoid the selection of primary nodes from being predicted in advance, this article proposes a primary node selection method based on verifiable random function (VRF). Moreover, a method is proposed of selecting transactions based on weighted round robin (WRR) scheduling to motivate nodes for the purpose of maintaining a high reputation score. Finally, the performance of the proposed scheme is verified by experiments whose results show that the proposed consensus mechanism shows better performance than other comparable schemes in terms of the time added to the blockchain, transaction per second (TPS), consensus success rate, and the time of obtaining transaction. And the proposed mechanism is less dependent on roadside unit (RSU).

Formal Security Analysis on dBFT Protocol of NEO

NEO is one of the top public chains worldwide. It adopts a new consensus algorithm called delegated Byzantine Fault Tolerance (dBFT). In this article, we formalize dBFT via the state machine replication model and point out its potential issues. Our theoretical analysis indicates that dBFT could guarantee neither liveness nor safety , even if the number of Byzantine nodes is no more than the threshold, which has contradicted the established security claim. Then, we identify two attacks and successfully simulate them. Finally, we provide recommendations. Notably, NEO official team has accepted our suggested fixes.

Resilient Consensus Control for Multi-Agent Systems: A Comparative Survey

Due to the openness of communication network and the complexity of system structures, multi-agent systems are vulnerable to malicious network attacks, which can cause intense instability to these systems. This article provides a survey of state-of-the-art results of network attacks on multi-agent systems. Recent advances on three types of attacks, i.e., those on DoS attacks, spoofing attacks and Byzantine attacks, the three main network attacks, are reviewed. Their attack mechanisms are introduced, and the attack model and the resilient consensus control structure are discussed, respectively, in detail, in terms of the theoretical innovation, the critical limitations and the change of the application. Moreover, some of the existing results along this line are given in a tutorial-like fashion. In the end, some challenges and open issues are indicated to guide future development directions of the resilient consensus of multi-agent system under network attacks.

Sharding-Based Proof-of-Stake Blockchain Protocols: Key Components & Probabilistic Security Analysis

Blockchain technology has been gaining great interest from a variety of sectors including healthcare, supply chain, and cryptocurrencies. However, Blockchain suffers from a limited ability to scale (i.e., low throughput and high latency). Several solutions have been proposed to tackle this. In particular, sharding has proved to be one of the most promising solutions to Blockchain's scalability issue. Sharding can be divided into two major categories: (1) Sharding-based Proof-of-Work (PoW) Blockchain protocols, and (2) Sharding-based Proof-of-Stake (PoS) Blockchain protocols. The two categories achieve good performances (i.e., good throughput with a reasonable latency), but raise security issues. This article focuses on the second category. In this paper, we start by introducing the key components of sharding-based PoS Blockchain protocols. We then briefly introduce two consensus mechanisms, namely PoS and practical Byzantine Fault Tolerance (pBFT), and discuss their use and limitations in the context of sharding-based Blockchain protocols. Next, we provide a probabilistic model to analyze the security of these protocols. More specifically, we compute the probability of committing a faulty block and measure the security by computing the number of years to fail. We achieve a number of years to fail of approximately 4000 in a network of 4000 nodes, 10 shards, and a shard resiliency of 33%.

Blockchain‐based reliable image copyright protection

AbstractThe indiscriminate distribution of a large number of unlicensed images through the Internet seriously harms the benefits of image owners, which prompts the study of image copyright protection solutions. Existing blockchain‐free solutions are limited to using centralized servers to store copyright information, leaving data vulnerable to being lost or tampered with by external attacks. Although the blockchain‐based image copyright protection solutions are capable of addressing the above problem, improvements are needed in choosing a suitable blockchain platform and supporting integrated copyright life cycle management. To this end, this paper designs a blockchain‐based reliable image copyright protection system named BB‐RICP using Hyperledger Fabric. For one thing, the adoption of Fabric mitigates the limitations of other blockchain platforms represented by Ethereum in terms of efficiency, economy, and availability. For another, BB‐RICP successes in achieving integrated copyright lifecycle management. In addition, BB‐RICP achieves enhanced applicability performance by firstly introducing GM algorithms recommended by Chinese national standard and adapting the more consortium blockchain‐oriented practical byzantine fault tolerance algorithm. Moreover, the adoption of spread spectrum watermarking brings convenience to users. Finally, Kubernetes is used to assist in the simulation of BB‐RICP to verify that users can operate reliably in the blockchain network.

A Taxonomic Hierarchy of Blockchain Consensus Algorithms: An Evolutionary Phylogeny Approach

Countless endeavors have been undertaken to address the Byzantine Generals Problem, a generalization of the Two Generals Problem. The emergence of proof of work (PoW) for Bitcoin has led to various consensus algorithms diverging, and comparable existing consensus algorithms are being gradually utilized interchangeably, or only developed for each specific application domain. Our approach employs an evolutionary phylogeny method to classify blockchain consensus algorithms based on their historical development and current usage. To demonstrate the relatedness and lineage of distinct algorithms, as well as to support the recapitulation theory, which posits that the evolutionary history of its mainnets is mirrored in the development of an individual consensus algorithm, we present a taxonomy. We have created a comprehensive classification of past and present consensus algorithms that serves to organize this swift consensus algorithm evolution period. By recognizing similarities, we have compiled a list of different verified consensus algorithms and performed clustering on over 38 of these. Our new taxonomic tree presents five taxonomic ranks, including the evolutionary process and decision-making method, as a technique for analyzing correlation. Through the examination of the evolution and utilization of these algorithms, we have developed a systematic and hierarchical taxonomy that enables the grouping of consensus algorithms into distinct categories. The proposed method classifies various consensus algorithms according to taxonomic ranks and aims to reveal the direction of research on the application of blockchain consensus algorithms for each domain.

WGM-dSAGA: Federated Learning Strategies with Byzantine Robustness Based on Weighted Geometric Median

Federated learning techniques accomplish federated modeling and share global models without sharing data. Federated learning offers a good answer to complex data and privacy security issues. Although there are many ways to target federated learning, Byzantine attacks are the ones we concentrate on. Byzantine attacks primarily impede learning by tampering with the local model parameters provided by a client to the master node throughout the federation learning process, leading to a final global model that diverges from the optimal solution. To address this problem, we combine aggregation rules with Byzantine robustness using a gradient descent optimization algorithm based on variance reduction. We propose a WGM-dSAGA method with Byzantine robustness, called weighted geometric median-based distributed SAGA. We replace the original mean aggregation strategy in the distributed SAGA with a robust aggregation rule based on weighted geometric median. When less than half of the clients experience Byzantine attacks, the experimental results demonstrate that our proposed WGM-dSAGA approach is highly robust to different Byzantine attacks. Our proposed WGM-dSAGA algorithm provides the optimal gap and variance under a Byzantine attack scenario.

The Byzantine Empire and Its Generals: An Ancient Empire Back to Life in Computer Security

Forty years after its initial publication, we revisit the seed contribution of Byzantine fault tolerance, focusing on its application for the security of systems implemented in software. We describe new environments in which it is being used.

Dynamic Secure Access Control and Data Sharing Through Trusted Delegation and Revocation in a Blockchain-Enabled Cloud-IoT Environment

The Internet of Things (IoT) is vulnerable to leakage of private information during data sharing. To avoid this problem, access control and secure data sharing have been introduced in IoT; however, many challenges are faced because of centralized access control and single delegator selection. Additionally, blockchain is integrated into IoT to enhance the security of the environment. For that purpose, this research proposes dynamic secure access control using the blockchain (DSA-Block) model, which performs secure access control and data sharing. Initially, the IoT device attributes and user attributes are registered at a local domain authority (LDA) for generating private and public keys using the hyperelliptic curve cryptography (HECC) algorithm, which ensures the legitimacy of the users and devices. Then, the IoT devices send a request message to the edge nodes (ENs) via a gateway, which performs request filtration by validating the user’s authenticity. The filtered requests are sent to the edge server to perform access delegation using rock hyraxes swarm optimization (RHSO), which selects a set of delegator nodes. The access control decision is made by using the Trusted practical Byzantine fault tolerance (PBFT) consensus algorithm. The IoT data are stored in the cloud server for secure storage, in which the data are secured using a differential privacy mechanism. Finally, dual revocations, such as user attribute revocation and user revocation, are used to maintain security. The performance of DSA-Block is evaluated and the results demonstrate that the proposed DSA-Block model achieves superior performance compared to previous works.

Novel Lightweight Consensus Protocol for Reaching Blockchain Consensus in Low Bandwidth Network Environment

As blockchain technology makes rapid progress, attempts to use it in various environments are also increasing. Especially home networks and smart factories consisting of a small network of Internet of Things (IoT) devices with singlehop or a few hops are trying to adopt blockchain systems to guarantee the integrity of data and to protect privacy. However, in such low-bandwidth network environments, they encounter challenges because of the network cost for block consensus. This study proposes a lightweight blockchain consensus protocol suitable for these environments. The proposed protocol improves the existing PBFT protocol to reduce the network cost required for consensus. The main innovation is to reduce the number of phases for consensus from three phases to two phases while preserving Byzantine failure tolerance. These two phases guarantee the total order of clients’ transactions in the same view. And the simple viewchange process in the proposed protocol enables the change of the primary node for new block generation. The experimental result shows a significant performance improvement of block consensus over the PBFT. Using the proposed protocol can contribute to operating blockchain systems in various lowbandwidth network environments. Keywords—blockchain, consensus protocol, lightweight protocol, low bandwidth network, PBFT

Enhancing the Security of Software Defined Mobile Networks (SDMN) based on Blockchain Technology

The blockchain involves future developments and new technologies. The emergence of a blockchain is a challenge for the conventional social organization and mode of activity. Data latency and mobile network capacity would no longer be a limitation for mobile users in next-generation networks of Software Defined Mobile Networks (SDMN). But there are many benefits to Software Defined Mobile Networking, it also contributes to certain security problems like DDoS / DoS attacks, unauthorized access, and single data point error. To enhance the security and privacy of the SDMN control plane, this paper proposes a new SDMN-based “Simplified Byzantine Fault Tolerance (SBFT)” algorithm to send signals between controllers and also set up an analysis study to investigate SBFT's security and results. However, there are many benefits of Software Defined Mobile Networking (SDMN), and then it helps to resolve other security concerns including DDoS / DoS attacks, unauthorized access, and single point failure. Blockchain, an evolving revolutionary technology, will offer creative approaches to address security issues in Software-Defined Mobile networks in an efficient way. This study proposes an SDMN-based Simplified Byzantine Fault Tolerance (SBFT) for enhancing the mobile network's security and privacy.

Enhanced Multiset Consensus Protocol Based on PBFT for Logistics Information Traceability

In the recent years, the global logistics industry has greatly driven the development of the world economy. At the same time, a large amount of data information is generated. Due to the frequent occurrence of logistics information leakage and forgery, it is necessary to find solutions that can accurately trace logistics information and ensure the security and authenticity of logistics information. The birth of blockchain technology has transformed the logistics industry from quantitative change to qualitative change. The technical characteristics of blockchain technology, such as distributed storage ideas, decentralization, immutability, and complex encryption consensus algorithm, endow it with a wide range of application prospects in the logistics industry. This paper proposes an enhanced multiset consensus algorithm based on PBFT (practical Byzantine fault tolerance) for logistics information traceability and storage on the logistics blockchain. The application of the proposed multi-set consensus algorithm in the topology structure composed of multiple sets can improve the consensus efficiency of logistics information in the blockchain. We improve consensus capability and transaction speed, avoid redundant consensus message packets occupying a large bandwidth, and efficiently process logistics information generated at any time. We ensure the traceability of logistics information and achieve efficient and accurate traceability, and the efficiency and security of the proposed algorithm are analyzed. This paper aims to solve the problems of traceability, trustworthiness, and efficient processing of blockchain applications in logistics information to operate the logistics network efficiently. This paper compares the proposed algorithm with the PBFT-related expansion algorithm regarding bandwidth occupation, delay, and throughput. The results show that the MPBFT consensus algorithm significantly improves the efficiency of the logistics blockchain network.

A decentralized crypto network with dynamic threshold change

AbstractThe threshold proxy re‐encryption (tPRE) algorithm not only ensures the separation of data ownership and use right, but also prevents a single data proxy from doing evil. In order to avoid collusion of multiple data proxies, this paper constructs a decentralized crypto network (DCN) based on blockchain technology and tPRE. DCN is more than a key management system, while expanding the data storage capacity of blockchain network. Combined with existing distributed storage schemes, it can store massive data. Due to Byzantine failure in distributed networks, the data proxy nodes can join or leave at will. In order to ensure that the number of data proxy nodes in the DCN network meets the requirements of the re‐encryption threshold, and the data authorization operation can be carried out smoothly, this paper presents a dynamic refresh method of tPRE fragments. When there are fragments that meet the threshold number in the network, the allocation of encrypted fragments can be adjusted non‐interactivity and self‐adaptively. Non‐interactivity means that each data proxy node independently refreshes its old fragments into new fragments without collecting old fragments from other nodes in the DCN, and then distributes the new fragments to other data proxy nodes in the DCN. Self‐adaptability refers to monitoring the changes of data proxy node sets in the DCN, and automatically refreshing fragments. Finally, the performance of critical algorithms is calculated in the proposal.

Total Variation-Based Distributed Kalman Filtering for Resiliency Against Byzantines

This article proposes a distributed Kalman filter (DKF) with enhanced robustness against Byzantine adversaries. A Byzantine agent is a legitimate network agent that, unlike an honest agent, manipulates information before sharing it with neighbors to impair the overall system performance. In contrast to the literature, the DKF is modeled as a distributed optimization problem where resiliency against Byzantine agents is accomplished by employing a total variation (TV) penalty term. We utilize a distributed subgradient algorithm to compute the state estimate and error covariance matrix updates of the DKF. Additionally, we prove that the proposed suboptimal solution converges to a neighborhood of the optimal centralized solution of the KF with a bounded radius when Byzantine agents are present. Numerical simulations corroborate the theoretical findings and demonstrate the robustness of the proposed DKF against Byzantine attacks.

Towards understanding quality challenges of the federated learning for neural networks: a first look from the lens of robustness

Federated learning (FL) is a distributed learning paradigm that preserves users’ data privacy while leveraging the entire dataset of all participants. In FL, multiple models are trained independently on the clients and aggregated centrally to update a global model in an iterative process. Although this approach is excellent at preserving privacy, FL still suffers from quality issues such as attacks or byzantine faults. Recent attempts have been made to address such quality challenges on the robust aggregation techniques for FL. However, the effectiveness of state-of-the-art (SOTA) robust FL techniques is still unclear and lacks a comprehensive study. Therefore, to better understand the current quality status and challenges of these SOTA FL techniques in the presence of attacks and faults, we perform a large-scale empirical study to investigate the SOTA FL’s quality from multiple angles of attacks, simulated faults (via mutation operators), and aggregation (defense) methods. In particular, we study FL’s performance on the image classification tasks and use Deep Neural Networks as our model type. Furthermore, we perform our study on two generic image datasets and one real-world federated medical image dataset. We also systematically investigate the effect of the proportion of affected clients and the dataset distribution factors on the robustness of FL. After a large-scale analysis with 496 configurations, we find that most mutators on each user have a negligible effect on the final model in the generic datasets, and only one of them is effective in the medical dataset. Furthermore, we show that model poisoning attacks are more effective than data poisoning attacks. Moreover, choosing the most robust FL aggregator depends on the attacks and datasets. Finally, we illustrate that a simple ensemble of aggregators achieves a more robust solution than any single aggregator and is the best choice in 75% of the cases. The data that support the findings of this study are available in our repository at https://github.com/aminesi/federated .

EBMICQL: Improving Efficiency of Blockchain Miner Pools via Incremental and Continuous Q-Learning Framework

Blockchain mining pools assist in reducing computational load on individual miner nodes via distributing mining tasks. This distribution must be done in a non-redundant manner, so that each miner is able to calculate block hashes with optimum efficiency. To perform this task, a wide variety of mining optimization methods are proposed by researchers, and most of them distribute mining tasks via statistical request processing models. These models segregate mining requests into non-redundant sets, each of which will be processed by individual miners. But this division of requests follows a static procedure, and does not consider miner specific parameters for set creation, due to which overall efficiency of the underlying model is limited, which reduces its mining performance under real-time scenarios. To overcome this issue, an Incremental & Continuous Q-Learning Framework for generation of miner-specific task groups is proposed in this text. The model initially uses a Genetic Algorithm (GA) method to improve individual miner performance, and then applies Q-Learning to individual mining requests. The Reason for selecting GA model is that it assists in maintaining better speed-to-power (S2P) ratio by optimization of miner resources that are utilized during computations. While, the reason for selecting Q-Learning Model is that it is able to continuously identify miners performance, and create performance-based mining pools at a per-miner level. Due to application of Q-Learning, the model is able to assign capability specific mining tasks to individual miner nodes. Because of this capability-driven approach, the model is able to maximize efficiency of mining, while maintaining its QoS performance. The model was tested on different consensus methods including Practical Byzantine Fault Tolerance Algorithm (PBFT), Proof-of-Work (PoW), Proof-of-Stake (PoS), and Delegated PoS (DPoS), and its performance was evaluated in terms of mining delay, miner efficiency, number of redundant calculations per miner, and energy efficiency for mining nodes. It was observed that the proposed GA based Q-Learning Model was able to reduce mining delay by 4.9%, improve miners efficiency by 7.4%, reduce number of redundant computations by 3.5%, and reduce energy required for mining by 7.1% when compared with various state-of-the-art mining optimization techniques. Similar performance improvement was observed when the model was applied on different blockchain deployments, thus indicating better scalability and deployment capability for multiple application scenarios.

SPDL: A Blockchain-Enabled Secure and Privacy-Preserving Decentralized Learning System

Decentralized learning involves training machine learning models over remote mobile devices, edge servers, or cloud servers while keeping data localized. Even though many studies have shown the feasibility of preserving privacy, enhancing training performance or introducing Byzantine resilience, but none of them simultaneously considers all of them. Therefore we face the following problem: <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">how can we efficiently coordinate the decentralized learning process while simultaneously maintaining learning security and data privacy for the entire system?</i> To address this issue, in this paper we propose SPDL, a blockchain-secured and privacy-preserving decentralized learning system. SPDL integrates blockchain, Byzantine Fault-Tolerant (BFT) consensus, BFT Gradients Aggregation Rule (GAR), and differential privacy seamlessly into one system, ensuring efficient machine learning while maintaining data privacy, Byzantine fault tolerance, transparency, and traceability. To validate our approach, we provide rigorous analysis on convergence and regret in the presence of Byzantine nodes. We also build a SPDL prototype and conduct extensive experiments to demonstrate that SPDL is effective and efficient with strong security and privacy guarantees.