FEEL: Federated End-to-End Learning With Non-IID Data for Vehicular Ad Hoc Networks

Li, Beibei; Jiang, Yukun; Pei, Qingqi; Li, Beibei; Liu, Liang; Lu, Rongxing

doi:10.1109/tits.2022.3190294

Cited by 17 publications

(4 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, it should be noted that the method of using selection or client-side clustering is essentially an asynchronous update mechanism. This method is widely utilized in standard FL, but experiments indicate that some methods directly implemented in ITS scenarios will decrease performance [127]. In addition, in dealing with the problem of uneven user resources, it is likewise possible to compensate for the performance loss of the stragglers or low-performance models via coding calculation, i.e., embedding calculation redundancy [128].…”

Section: Generated Dataset 2021mentioning

confidence: 99%

“…Regarding that the single-hop cluster structure limits the system's coverage and stability, numerous clustering algorithms based on passive multi-hop have been developed in recent years [137], [138]. Using the network topology of ITS, the work of [127] designs an end-to-end FL framework, as well as inter-cluster and inner-cluster learning algorithms, which accurately reduce redundant communication overhead. The above end-to-end approach, which can direct vehicles directly from raw data without costly labels, is widely used in ADS at present.…”

Section: Generated Dataset 2021mentioning

confidence: 99%

See 1 more Smart Citation

Enhanced Federated Learning for Edge Data Security in Intelligent Transportation Systems

Zhu

Yin

et al. 2023

IEEE Trans. Intell. Transport. Syst.

View full text Add to dashboard Cite

Section: Generated Dataset 2021mentioning

confidence: 99%

Section: Generated Dataset 2021mentioning

confidence: 99%

Enhanced Federated Learning for Edge Data Security in Intelligent Transportation Systems

Zhu

Yin

et al. 2023

IEEE Trans. Intell. Transport. Syst.

View full text Add to dashboard Cite

“…Ma et al adaptively adjusted the learning rate of the local model based on the frequency of each client's participation in global training, thus improving the accuracy in the Non-IID case [27]. Li et al designed a weighted inter-cluster recurrent update algorithm and an internal cluster recursive learning method to reduce redundant communication costs and improve the performance of deep learning models for Non-IID data [28]. Shu et al adaptively adjust the threshold to filter the training data for each client by capturing the different data distributions among clients, and selecting the suitable clients to participate in each round of global learning thus can reduce the bias of the weights and improve the model accuracy [29].…”

Section: Non-iid Problem Of Flmentioning

confidence: 99%

FedSiM: a similarity metric federal learning mechanism based on stimulus response method with Non-IID data

Wang,

Zhang

2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

FL (Federal Learning) based on parameter sharing under the assumption that the data obey IID (Independent Identical Distribution) has already achieved good results in areas such as fault diagnosis. Data collected by the decentralized devices often do not obey IID. However, when faced with the scenario of client data obeying Non-IID distribution, its diagnostic accuracy is usually weak. Based on this, we did an investigation on the mechanism causing this phenomenon and found that it was attributed to the weight shift of the network. Therefore, based on the stimulus response principle, we investigated the network similarity of federal clients under different data distributions and explain the reasons for the weight shift. Firstly, it was pointed out that there are differences in the regions where the network is activated when performing different classification tasks. Then, FedSiM (Similarity Metric Federal Learning) was proposed based on the principle that there are differences between the activated regions. Finally, experiments were designed on the Case Western Reserve University bearing failure dataset for different degrees of independent identical distribution cases. The results show that FedSiM can improve the diagnostic accuracy by 15.8 percentage points in the case of Non-IID, and a few shared FedSiM methods to further improve the accuracy were also given.

show abstract

“…The fluctuating distribution of training data held by passing vehicles hinders traditional learning methods from achieving optimal performance. It is a significant challenge to achieve qualified learning on the non-IID data of vehicles by decentralized federated learning [17].…”

Section: Introductionmentioning

confidence: 99%

Hierarchical Decentralized Federated Learning Framework with Adaptive Clustering: Bloom-Filter-Based Companions Choice for Learning Non-IID Data in IoV

Liu,

et al. 2023

Electronics

View full text Add to dashboard Cite

The accelerating progress of the Internet of Vehicles (IoV) has put forward a higher demand for distributed model training and data sharing in vehicular networks. Traditional centralized approaches are no longer applicable in the face of drivers’ concerns about data privacy, while Decentralized Federated Learning (DFL) provides new possibilities to address this issue. However, DFL still faces challenges regarding the non-IID data of passing vehicles. To tackle this challenge, a novel DFL framework, Hierarchical Decentralized Federated Learning (H-DFL), is proposed to achieve qualified distributed training among vehicles by considering data complementarity. We include vehicles, base stations, and data center servers in this framework. Firstly, a novel vehicle-clustering paradigm is designed to group passing vehicles based on the Bloom-filter-based compact representation of data complementarity. In this way, vehicles train their models based on local data, exchange model parameters in each group, and achieve a qualified local model without the interference of imbalanced data. On a higher level, a local model trained by each group is submitted to the data center to obtain a model covering global features. Base stations maintain the local models of different groups and judge whether the local models need to be updated according to the global model. The experimental results based on real-world data demonstrate that H-DFL dose not only reduces communication latency with different participants but also addresses the challenges of non-IID data in vehicles.

show abstract

FEEL: Federated End-to-End Learning With Non-IID Data for Vehicular Ad Hoc Networks

Cited by 17 publications

References 36 publications

Enhanced Federated Learning for Edge Data Security in Intelligent Transportation Systems

Enhanced Federated Learning for Edge Data Security in Intelligent Transportation Systems

FedSiM: a similarity metric federal learning mechanism based on stimulus response method with Non-IID data

Hierarchical Decentralized Federated Learning Framework with Adaptive Clustering: Bloom-Filter-Based Companions Choice for Learning Non-IID Data in IoV

Contact Info

Product

Resources

About