Blockchain Assisted Decentralized Federated Learning (BLADE-FL): Performance Analysis and Resource Allocation

Li, Jun; Shao, Yumeng; Wei, Kang; Ding, Ming; Ma, Chuan; Shi, Long; Han, Zhu; Poor, H. Vincent

doi:10.1109/tpds.2021.3138848

Cited by 115 publications

(24 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since clients should ensure that the rewards they receive are not less than the total costs they spend, in such a situation, they can participate in the BCFL task. So in the situation of incomplete information, the MO needs to guarantee that its decisions should satisfy (17) to encourage clients to join the work. Besides, µ i of client i is not known by the MO, and the decisions of the MO are required to be based on the correct value of µ i reported by clients, so the MO needs to satisfy (18) when making the decisions.…”

Section: Resource Allocation With Incomplete In-formationmentioning

confidence: 99%

“…In [17], the resource allocation problem is resolved for the local devices with the same computational power in BCFL. An upper-bound of the global loss function was proposed to evaluate the performance of training; in the meantime, the relationship among update rounds, block generation rate, and learning rate was explored.…”

Section: Resource Allocation In Bcflmentioning

confidence: 99%

“…There exist very few studies that tackle the above challenges [17], [18]. The existing studies are based on two assumptions which are not practical: 1) all clients have identical computational power and data volume; and 2) the system knows all the information about clients.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Incentive Mechanism Design for Joint Resource Allocation in Blockchain-based Federated Learning

Wang¹,

Hu²,

Li³

et al. 2022

Preprint

View full text Add to dashboard Cite

Blockchain-based federated learning (BCFL) has recently gained tremendous attention because of its advantages such as decentralization and privacy protection of raw data. However, there has been few research focusing on the allocation of resources for clients in BCFL. In the BCFL framework where the FL clients and the blockchain miners are the same devices, clients broadcast the trained model updates to the blockchain network and then perform mining to generate new blocks. Since each client has a limited amount of computing resources, the problem of allocating computing resources into training and mining needs to be carefully addressed. In this paper, we design an incentive mechanism to assign each client appropriate rewards for training and mining, and then the client will determine the amount of computing power to allocate for each subtask based on these rewards using the two-stage Stackelberg game. After analyzing the utilities of the model owner (MO) (i.e., the BCFL task publisher) and clients, we transform the game model into two optimization problems, which are sequentially solved to derive the optimal strategies for both the MO and clients. Further, considering the fact that local training related information of each client may not be known by others, we extend the game model with analytical solutions to the incomplete information scenario. Extensive experimental results demonstrate the validity of our proposed schemes.

show abstract

Section: Resource Allocation With Incomplete In-formationmentioning

confidence: 99%

Section: Resource Allocation In Bcflmentioning

confidence: 99%

See 1 more Smart Citation

Incentive Mechanism Design for Joint Resource Allocation in Blockchain-based Federated Learning

Wang¹,

Hu²,

Li³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Ramanan et al [30] leverage Smart Contracts (SC) to coordinate the round delineation, model aggregation, and update tasks in FL. Li et al [17] allow each client to broadcast the trained model to other clients, to aggregate its own model with received ones, and then to compete to generate a block before its local training of the next round. Li et al [19] use blockchain for global model storage and local model update exchange and devise an innovative committee consensus mechanism to reduce consensus computing and malicious attacks.…”

Section: Related Workmentioning

confidence: 99%

“…One path is on-device training with no raw data or intermediate results uploading [10,47], which suffers the over-fitting problem. The other path is decentralized FL [14,15,17,19,30,39,40], a new FL paradigm to leverage a blockchain to coordinate the model aggregation and update parameters in a decentralized manner. Swarm Learning (SL) is the most representative and state-of-the-art decentralized FL paradigm [40], which combines decentralized hardware infrastructures, distributed machine learning with a permissioned blockchain to securely onboard members, dynamically elect the leader, and merge model parameters.…”

Section: Introductionmentioning

confidence: 99%

Demystifying Swarm Learning: A New Paradigm of Blockchain-based Decentralized Federated Learning

Han¹,

Ma²,

Han³

2022

Preprint

View full text Add to dashboard Cite

Federated learning (FL) is an emerging promising privacy-preserving machine learning paradigm and has raised more and more attention from researchers and developers. FL keeps users' private data on devices and exchanges the gradients of local models to cooperatively train a shared Deep Learning (DL) model on central custodians. However, the security and fault tolerance of FL have been increasingly discussed, because its central custodian mechanism or star-shaped architecture can be vulnerable to malicious attacks or software failures. To address these problems, Swarm Learning (SL) introduces a permissioned blockchain to securely onboard members and dynamically elect the leader, which allows performing DL in an extremely decentralized manner. Compared with tremendous attention to SL, there are few empirical studies on SL or blockchain-based decentralized FL, which provide comprehensive knowledge of best practices and precautions of deploying SL in real-world scenarios. Therefore, we conduct the first comprehensive study of SL to date, to fill the knowledge gap between SL deployment and developers, as far as we are concerned. In this paper, we conduct various experiments on 3 public datasets of 5 research questions, present interesting findings, quantitatively analyze the reasons behind these findings, and provide developers and researchers with practical suggestions. The findings have evidenced that SL is supposed to be suitable for most application scenarios, no matter whether the dataset is balanced, polluted, or biased over irrelevant features.

show abstract

An intelligent impulsive noise mitigation with deep learning method

Yang,

Qian,

Wang

et al. 2024

Int Journal of Mech Sys Dyn

View full text Add to dashboard Cite

To enable message transmission among sensors and equipment, power line communication (PLC) is a widely adopted smart grid. However, due to the occurrence of impulsive noise (IN), reliable transmissions over PLC channels in the smart grid are challenging. Therefore, in this paper, we propose an adaptive noise mitigation scheme to clip the IN with the sliding window‐based method, where the altitude of the received signal in the current time slots is obtained by computing the average altitude of signals in the previous and next time slots. To detect the states of IN and dynamically estimate the power threshold of signals for the IN mitigation scheme, we develop an intelligent algorithm based on the long short‐term memory network. To prevent the useful signals from being eliminated as IN signals, we propose the accelerated proximal gradient method (APGM) based on tone reservation to reduce the peak‐to‐average power ratio (PAPR) for the transmitting signals with low computational complexity. In addition, the closed‐form expression of the bit error rate (BER) is derived for the proposed sliding window‐based IN mitigation scheme according to the probability density function of the IN. Simulation results demonstrate that the proposed IN mitigation scheme achieves a better BER performance than the conventional IN mitigation schemes. In addition, the APGM aided by IN mitigation can further improve BER performance due to the PAPR reduction.

show abstract

Blockchain Assisted Decentralized Federated Learning (BLADE-FL): Performance Analysis and Resource Allocation

Cited by 115 publications

References 46 publications

Incentive Mechanism Design for Joint Resource Allocation in Blockchain-based Federated Learning

Incentive Mechanism Design for Joint Resource Allocation in Blockchain-based Federated Learning

Demystifying Swarm Learning: A New Paradigm of Blockchain-based Decentralized Federated Learning

An intelligent impulsive noise mitigation with deep learning method

Contact Info

Product

Resources

About