Federated Learning Model Research Articles

AUCTION: Automated and Quality-Aware Client Selection Framework for Efficient Federated Learning

The emergency of federated learning (FL) enables distributed data owners to collaboratively build a global model without sharing their raw data, which creates a new business chance for building data market. However, in practical FL scenarios, the hardware conditions and data resources of the participant clients can vary significantly, leading to different positive/negative effects on the FL performance, where the client selection problem becomes crucial. To this end, we propose <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AUCTION</i> , an <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A</u> utomated and q <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">U</u> ality-aware <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</u> lient selec <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TION</u> framework for efficient FL, which can evaluate the learning quality of clients and select them automatically with quality-awareness for a given FL task within a limited budget. To design <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AUCTION</i> , multiple factors such as data size, data quality, and learning budget that can affect the learning performance should be properly balanced. It is nontrivial since their impacts on the FL model are intricate and unquantifiable. Therefore, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AUCTION</i> is designed to encode the client selection policy into a neural network and employ reinforcement learning to automatically learn client selection policies based on the observed client status and feedback rewards quantified by the federated learning performance. In particular, the policy network is built upon an encoder-decoder deep neural network with an attention mechanism, which can adapt to dynamic changes of the number of candidate clients and make sequential client selection actions to reduce the learning space significantly. Extensive experiments are carried out based on real-world datasets and well-known learning models to demonstrate the efficiency, robustness, and scalability of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AUCTION</i> .

Joint Scheduling and Resource Allocation for Hierarchical Federated Edge Learning

The concept of hierarchical federated edge learning (H-FEEL) has been recently proposed as an enhancement of federated learning model. Such a system generally consists of three entities, i.e., the server, helpers, and clients, in which each helper collects the trained gradients from clients nearby, aggregates them, and sends the result to the server for global model update. Due to limited communication resources, only a portion of helpers can be scheduled to upload their aggregated gradients in each round of the model training. And that necessitates a well-designed scheme for the joint helper scheduling and communication resources allocation. In this paper, we develop a training algorithm for the H-FEEL system which involves local gradient computing, weighted gradient uploading, and machine learning model updating phases. By characterizing these phases mathematically and analyzing one-round convergence bound of the training algorithm, we formulate an optimization problem to achieve the scheduling and resource allocation scheme. The problem simultaneously captures the uncertainty of the wireless channel and the importance of the weighted gradient. To solve the problem, we first transform it into an equivalent problem and then decompose the transformed problem into two subproblems: <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">bit and sub-channel allocation</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">helper scheduling</i> , which are mixed integer nonlinear programming and continuous nonlinear problems, respectively. For the first subproblem, we obtain an optimal solution of exponential complexity and a suboptimal solution that has polynomial complexity. For the second subproblem, we obtain a closed-form optimal solution in a special case and a suboptimal solution in the general case. The efficacy of our scheme is amply demonstrated via simulations and the analytical framework is shown to provide valuable design insights for the practical implementation of the H-FEEL system.

WiFederated: Scalable WiFi Sensing Using Edge-Based Federated Learning

WiFi sensing using channel state information (CSI) offers a device-free and nonintrusive method for human activity monitoring. However, the data-hungry and location-specific training process hinders its scalable deployment at large sizes. In this work, we propose <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">WiFederated</i> , a federated learning (FL) approach to train machine learning models for WiFi sensing tasks. Using WiFederated, client devices can not only perform training in parallel at the edge instead of sequentially at a central server but can also collaboratively learn and share generalizable location-independent traits about physical actions being monitored. We demonstrate that an FL model trained on as few as 2–3 locations can provide high prediction accuracy in new locations even without any data available from them. We also demonstrate how new locations can achieve higher prediction accuracy even with a small number of available samples when using the pretrained FL model rather than training from scratch. The results show that the FL model can save local training epochs and reduce the need for large data collection at each new location. Thus, the proposed WiFederated system scales as more locations are added. We show that WiFederated provides a more accurate and time-efficient solution compared to existing transfer learning and adversarial learning solutions thanks to the parallel training ability at multiple clients. By introducing new client selection methods during the FL process, we also show that accuracy can further increase. Finally, we evaluate the feasibility of training models at the edge and introduce continuous annotation to allow for continuous learning over time.

Membership Inference Defense in Distributed Federated Learning Based on Gradient Differential Privacy and Trust Domain Division Mechanisms

Distributed federated learning models are vulnerable to membership inference attacks (MIA) because they remember information about their training data. Through a comprehensive privacy analysis of distributed federated learning models, we design an attack model based on generative adversarial networks (GAN) and member inference attacks (MIA). Malicious participants (attackers) utilize the attack model to successfully reconstruct training sets of other regular participants without any negative impact on the global model. To solve this problem, we apply the differential privacy method to the training process of the model, which effectively reduces the accuracy of member inference attacks by clipping the gradient and adding noise to it. In addition, we manage the participants hierarchically through the method of trust domain division to alleviate the performance degradation of the model caused by differential privacy processing. Experimental results show that in distributed federated learning, our designed scheme can effectively defend against member inference attacks in white-box scenarios and maintain the usability of the global model, realizing an effective trade-off between privacy and usability.

6G enabled federated learning for secure IoMT resource recommendation and propagation analysis

The addition of devices in the IoT framework has leveraged the performance and ability of medical device computation in developing the Internet of Medical Things (IoMT) framework. The 6G ecosystem needs to be redefined in comparison with earlier generation communication environment and setup. This article proposes a novel technique of resource recommendation and scheduling-based propagation analysis usingFederated Learning (FL). The technique is driven by user-log record analysis and extraction of similar patterns. The logs are further evaluated with an available spectrum and bandwidth ratio of 6G to compute the requirement and availability of resources. The aim of the federated learning model is based on resource patterns and resource event occurrence fetched from computational logs of user participation. The framework uses the resource-attribute optimization technique for customization. The recommendation is based on strategic evaluation and the Dynamic User Allocation (DUA) technique. The approach has integrated data evaluation from the Operating system, networking channel, and communication devices of IoMT and has compared performance over the standard resource recommendation model. The proposed technique has achieved 94.72% of accuracy over standard DUA datasets.

Efficient Privacy-Preserving Federated Learning With Unreliable Users

Federated learning (FL) has emerged as a powerful technology widely applied in Internet of Things (IoT). Recently, researchers have shown an increased interest in privacy-preserving FL with <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">unreliable users</i> . The goal of such works is to achieve private training under ciphertext mode while ensuring that the FL model is mainly derived from the contributions of users with high-quality data. However, the existing work is still in its infancy, and the main challenge faced by many researchers is how to achieve their schemes for meeting the demands of high accuracy and efficiency. To combat that, we propose an efficient privacy-preserving FL (EPPFL) scheme with <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">unreliable users</i> . Specifically, we design a novel scheme to mitigate the negative impact of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">unreliable users</i> , where the targeted model is guaranteed to be updated with high-quality data. Through iteratively executing our “Excluding Irrelevant Components” and “Weighted Aggregation,” the FL model converges rapidly while taking limited communication and computation overhead. As a result, not only the model accuracy can be optimized, but also the training efficiency can be improved. Meanwhile, we conduct a secure framework based on the threshold Paillier cryptosystem, which can rigorously protect all user-related private information during the training process. Furthermore, the extensive experiments demonstrate our EPPFL with high-level performance in terms of accuracy and efficiency.

Perturbation-enabled Deep Federated Learning for Preserving Internet of Things-based Social Networks

Federated Learning (FL), as an emerging form of distributed machine learning (ML), can protect participants’ private data from being substantially disclosed to cyber adversaries. It has potential uses in many large-scale, data-rich environments, such as the Internet of Things (IoT), Industrial IoT, Social Media (SM), and the emerging SM 3.0. However, federated learning is susceptible to some forms of data leakage through model inversion attacks. Such attacks occur through the analysis of participants’ uploaded model updates. Model inversion attacks can reveal private data and potentially undermine some critical reasons for employing federated learning paradigms. This article proposes novel differential privacy (DP)-based deep federated learning framework. We theoretically prove that our framework can fulfill DP’s requirements under distinct privacy levels by appropriately adjusting scaled variances of Gaussian noise. We then develop a Differentially Private Data-Level Perturbation (DP-DLP) mechanism to conceal any single data point’s impact on the training phase. Experiments on real-world datasets, specifically the social media 3.0, Iris, and Human Activity Recognition (HAR) datasets, demonstrate that the proposed mechanism can offer high privacy, enhanced utility, and elevated efficiency. Consequently, it simplifies the development of various DP-based FL models with different tradeoff preferences on data utility and privacy levels.

A Multi-Agent Reinforcement Learning Approach for Efficient Client Selection in Federated Learning

Federated learning (FL) is a training technique that enables client devices to jointly learn a shared model by aggregating locally computed models without exposing their raw data. While most of the existing work focuses on improving the FL model accuracy, in this paper, we focus on the improving the training efficiency, which is often a hurdle for adopting FL in real world applications. Specifically, we design an efficient FL framework which jointly optimizes model accuracy, processing latency and communication efficiency, all of which are primary design considerations for real implementation of FL. Inspired by the recent success of Multi Agent Reinforcement Learning (MARL) in solving complex control problems, we present FedMarl, a federated learning framework that relies on trained MARL agents to perform efficient run-time client selection. Experiments show that FedMarl can significantly improve model accuracy with much lower processing latency and communication cost.

CrowdFL: A Marketplace for Crowdsourced Federated Learning

Amid data privacy concerns, Federated Learning (FL) has emerged as a promising machine learning paradigm that enables privacy-preserving collaborative model training. However, there exists a need for a platform that matches data owners (supply) with model requesters (demand). In this paper, we present CrowdFL, a platform to facilitate the crowdsourcing of FL model training. It coordinates client selection, model training, and reputation management, which are essential steps for the FL crowdsourcing operations. By implementing model training on actual mobile devices, we demonstrate that the platform improves model performance and training efficiency. To the best of our knowledge, it is the first platform to support crowdsourcing-based FL on edge devices.

Contribution-Aware Federated Learning for Smart Healthcare

Artificial intelligence (AI) is a promising technology to transform the healthcare industry. Due to the highly sensitive nature of patient data, federated learning (FL) is often leveraged to build models for smart healthcare applications. Existing deployed FL frameworks cannot address the key issues of varying data quality and heterogeneous data distributions across multiple institutions in this sector. In this paper, we report our experience developing and deploying the Contribution-Aware Federated Learning (CAFL) framework for smart healthcare. It provides an efficient and accurate approach to fairly evaluate FL participants' contribution to model performance without exposing their private data, and improves the FL model training protocol to allow the best performing intermediate models to be distributed to participants for FL training. Since its deployment in Yidu Cloud Technology Inc. in March 2021, CAFL has served 8 well-established medical institutions in China to build healthcare decision support models. It can perform contribution evaluations 2.84 times faster than the best existing approach, and has improved the average accuracy of the resulting models by 2.62% compared to the previous system (which is significant in industrial settings). To our knowledge, it is the first contribution-aware federated learning successfully deployed in the healthcare industry.

FedInv: Byzantine-Robust Federated Learning by Inversing Local Model Updates

Federated learning (FL) is a privacy-preserving distributed machine learning paradigm that enables multiple clients to collaboratively train statistical models without disclosing raw training data. However, the inaccessible local training data and uninspectable local training process make FL susceptible to various Byzantine attacks (e.g., data poisoning and model poisoning attacks), aiming to manipulate the FL model training process and degrade the model performance. Most of the existing Byzantine-robust FL schemes cannot effectively defend against stealthy poisoning attacks that craft poisoned models statistically similar to benign models. Things worsen when many clients are compromised or data among clients are highly non-independent and identically distributed (non-IID). In this work, to address these issues, we propose FedInv, a novel Byzantine-robust FL framework by inversing local model updates. Specifically, in each round of local model aggregation in FedInv, the parameter server first inverses the local model updates submitted by each client to generate a corresponding dummy dataset. Then, the server identifies those dummy datasets with exceptional Wasserstein distances from others and excludes the related local model updates from model aggregation. We conduct an exhaustive experimental evaluation of FedInv. The results demonstrate that FedInv significantly outperforms the existing robust FL schemes in defending against stealthy poisoning attacks under highly non-IID data partitions.

A federated learning model based on filtering strategy

A federated learning model based on filtering strategy

FLED-Block: Federated Learning Ensembled Deep Learning Blockchain Model for COVID-19 Prediction.

With the SARS-CoV-2's exponential growth, intelligent and constructive practice is required to diagnose the COVID-19. The rapid spread of the virus and the shortage of reliable testing models are considered major issues in detecting COVID-19. This problem remains the peak burden for clinicians. With the advent of artificial intelligence (AI) in image processing, the burden of diagnosing the COVID-19 cases has been reduced to acceptable thresholds. But traditional AI techniques often require centralized data storage and training for the predictive model development which increases the computational complexity. The real-world challenge is to exchange data globally across hospitals while also taking into account of the organizations' privacy concerns. Collaborative model development and privacy protection are critical considerations while training a global deep learning model. To address these challenges, this paper proposes a novel framework based on blockchain and the federated learning model. The federated learning model takes care of reduced complexity, and blockchain helps in distributed data with privacy maintained. More precisely, the proposed federated learning ensembled deep five learning blockchain model (FLED-Block) framework collects the data from the different medical healthcare centers, develops the model with the hybrid capsule learning network, and performs the prediction accurately, while preserving the privacy and shares among authorized persons. Extensive experimentation has been carried out using the lung CT images and compared the performance of the proposed model with the existing VGG-16 and 19, Alexnets, Resnets-50 and 100, Inception V3, Densenets-121, 119, and 150, Mobilenets, SegCaps in terms of accuracy (98.2%), precision (97.3%), recall (96.5%), specificity (33.5%), and F1-score (97%) in predicting the COVID-19 with effectively preserving the privacy of the data among the heterogeneous users.

PEPFL: A framework for a practical and efficient privacy-preserving federated learning

As an emerging joint learning model, federated learning is a promising way to combine model parameters of different users for training and inference without collecting users’ original data. However, a practical and efficient solution has not been established in previous work due to the absence of efficient matrix computation and cryptography schemes in the privacy-preserving federated learning model, especially in partially homomorphic cryptosystems. In this paper, we propose a Practical and Efficient Privacy-preserving Federated Learning (PEPFL) framework. First, we present a lifted distributed ElGamal cryptosystem for federated learning, which can solve the multi-key problem in federated learning. Secondly, we develop a Practical Partially Single Instruction Multiple Data (PSIMD) parallelism scheme that can encode a plaintext matrix into single plaintext for encryption, improving the encryption efficiency and reducing the communication cost in partially homomorphic cryptosystem. In addition, based on the Convolutional Neural Network (CNN) and the designed cryptosystem, a novel privacy-preserving federated learning framework is designed by using Momentum Gradient Descent (MGD). Finally, we evaluate the security and performance of PEPFL. The experiment results demonstrate that the scheme is practicable, effective, and secure with low communication and computation costs.

A blockchain based privacy-preserving federated learning scheme for Internet of Vehicles

The application of artificial intelligence technology in Internet of Vehicles (IoV) has attracted great research interests with the goal of enabling smart transportation and traffic management. Meanwhile, concerns have been raised over the security and privacy of the tons of traffic and vehicle data. In this regard, Federated Learning (FL) with privacy protection features is considered a highly promising solution. However, in the FL process, the server side may take advantage of its dominant role in model aggregation to steal sensitive information of users, while the client side may also upload malicious data to compromise the training of the global model. Most existing privacy-preserving FL schemes in IoV fail to deal with threats from both of these two sides at the same time. In this paper, we propose a Blockchain based Privacy-preserving Federated Learning scheme named BPFL, which uses blockchain as the underlying distributed framework of FL. We improve the Multi-Krum technology and combine it with the homomorphic encryption to achieve ciphertext-level model aggregation and model filtering, which can enable the verifiability of the local models while achieving privacy-preservation. Additionally, we develop a reputation-based incentive mechanism to encourage users in IoV to actively participate in the federated learning and to practice honesty. The security analysis and performance evaluations are conducted to show that the proposed scheme can meet the security requirements and improve the performance of the FL model.

Food Risk Entropy Model Based on Federated Learning

The safety of agricultural products is a guarantee of national security. The increasing variety of pesticides used on crops has led to an increasing abundance of pesticide residues in agricultural products, making pesticide residues an important factor in threatening health. Traditional indicators for evaluating the safety of agricultural products, such as pass rates and residue rates, can only qualitatively describe the level of pesticide residues. Isolated data leads to low data utilization, data is distributed between different terminals or departments and cannot be shared, while the security of private data needs to be ensured. Therefore, we propose a risk entropy model based on federated learning. The model is able to quantitatively describe the risk level of agricultural products and achieve data fusion without exposing private data in the federated learning model. In this paper, a total of 90,510 agricultural product data samples from 2015 to 2019 are collected, with each sample containing 58 indicators. The experimental results show that the developed food safety risk entropy model can quantitatively reflect the level of risk in the target region and time interval. In addition, we have developed a multidimensional data analysis tool based on federated learning, which can achieve data integration across multiple regions and departments.

FedSGDCOVID: Federated SGD COVID-19 Detection under Local Differential Privacy Using Chest X-ray Images and Symptom Information.

Coronavirus (COVID-19) has created an unprecedented global crisis because of its detrimental effect on the global economy and health. COVID-19 cases have been rapidly increasing, with no sign of stopping. As a result, test kits and accurate detection models are in short supply. Early identification of COVID-19 patients will help decrease the infection rate. Thus, developing an automatic algorithm that enables the early detection of COVID-19 is essential. Moreover, patient data are sensitive, and they must be protected to prevent malicious attackers from revealing information through model updates and reconstruction. In this study, we presented a higher privacy-preserving federated learning system for COVID-19 detection without sharing data among data owners. First, we constructed a federated learning system using chest X-ray images and symptom information. The purpose is to develop a decentralized model across multiple hospitals without sharing data. We found that adding the spatial pyramid pooling to a 2D convolutional neural network improves the accuracy of chest X-ray images. Second, we explored that the accuracy of federated learning for COVID-19 identification reduces significantly for non-independent and identically distributed (Non-IID) data. We then proposed a strategy to improve the model’s accuracy on Non-IID data by increasing the total number of clients, parallelism (client-fraction), and computation per client. Finally, for our federated learning model, we applied a differential privacy stochastic gradient descent (DP-SGD) to improve the privacy of patient data. We also proposed a strategy to maintain the robustness of federated learning to ensure the security and accuracy of the model.

A state-of-the-art survey on solving non-IID data in Federated Learning

Federated Learning (FL) proposed in recent years has received significant attention from researchers in that it can enable multiple clients to cooperatively train global models without revealing private data. This training mode protects users’ privacy without violating the supervision, and aggregates scattered data to exert great potential. However, the data samples on each participating device of FL are usually not independent and identically distributed (IID), which leads to serious statistical heterogeneity challenges for FL. In this article, we analyze and establish the definition of non-IID data problems, and put forward a series of challenges that this problem may bring to FL. We classify existing methods to solve this problem from the researcher’s entry point and subsequent sub-methods, aiming to provide a comprehensive study for solving the problem of non-IID data in FL. Our research shows that non-IID data will not only reduce the performance of the FL model, but also damage the active participation of users in the FL process. Compared with methods based on data-side sharing, enhancement, and selection, it is more common for researchers to improve federated learning algorithms from models, algorithms, and frameworks to solve non-IID problems. To the best of our knowledge, although many efforts have been made to address the problem of non-IID data, there are currently few authoritative systematic reviews in this field and are not up-to-date. In this article, we will fill the gaps in FL and provide researchers with the state-of-the-art research results to solve non-IID problems in FL and promote the further implementation of FL.

Architectural patterns for the design of federated learning systems

Federated learning has received fast-growing interests from academia and industry to tackle the challenges of data hungriness and privacy in machine learning. A federated learning system can be viewed as a large-scale distributed system with different components and stakeholders as numerous client devices participate in federated learning. Designing a federated learning system requires software system design thinking apart from the machine learning knowledge. Although much effort has been put into federated learning from the machine learning technique aspects, the software architecture design concerns in building federated learning systems have been largely ignored. Therefore, in this paper, we present a collection of architectural patterns to deal with the design challenges of federated learning systems. Architectural patterns present reusable solutions to a commonly occurring problem within a given context during software architecture design. The presented patterns are based on the results of a systematic literature review and include three client management patterns, four model management patterns, three model training patterns, four model aggregation patterns, and one configuration pattern. The patterns are associated to the particular state transitions in a federated learning model lifecycle, serving as a guidance for effective use of the patterns in the design of federated learning systems.

GTG-Shapley: Efficient and Accurate Participant Contribution Evaluation in Federated Learning

Federated Learning (FL) bridges the gap between collaborative machine learning and preserving data privacy. To sustain the long-term operation of an FL ecosystem, it is important to attract high-quality data owners with appropriate incentive schemes. As an important building block of such incentive schemes, it is essential to fairly evaluate participants’ contribution to the performance of the final FL model without exposing their private data. Shapley Value (SV)–based techniques have been widely adopted to provide a fair evaluation of FL participant contributions. However, existing approaches incur significant computation costs, making them difficult to apply in practice. In this article, we propose the Guided Truncation Gradient Shapley (GTG-Shapley) approach to address this challenge. It reconstructs FL models from gradient updates for SV calculation instead of repeatedly training with different combinations of FL participants. In addition, we design a guided Monte Carlo sampling approach combined with within-round and between-round truncation to further reduce the number of model reconstructions and evaluations required. This is accomplished through extensive experiments under diverse realistic data distribution settings. The results demonstrate that GTG-Shapley can closely approximate actual Shapley values while significantly increasing computational efficiency compared with the state-of-the-art, especially under non-i.i.d. settings.