Personalized Fair Split Learning for Resource-Constrained Internet of Things
With the flourishing development of the Internet of Things (IoT), federated learning has garnered significant attention as a distributed learning method aimed at preserving the privacy of participant data. However, certain IoT devices, such as sensors, face challenges in effectively employing conven...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-12-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/24/1/88 |
| _version_ | 1827384467997589504 |
|---|---|
| author | Haitian Chen Xuebin Chen Lulu Peng Yuntian Bai |
| author_facet | Haitian Chen Xuebin Chen Lulu Peng Yuntian Bai |
| author_sort | Haitian Chen |
| collection | DOAJ |
| description | With the flourishing development of the Internet of Things (IoT), federated learning has garnered significant attention as a distributed learning method aimed at preserving the privacy of participant data. However, certain IoT devices, such as sensors, face challenges in effectively employing conventional federated learning approaches due to limited computational and storage resources, which hinder their ability to train complex local models. Additionally, in IoT environments, devices often face problems of data heterogeneity and uneven benefit distribution between them. To address these challenges, a personalized and fair split learning framework is proposed for resource-constrained clients. This framework first adopts a U-shaped structure, dividing the model to enable resource-constrained clients to offload subsets of the foundational model to a central server while retaining personalized model subsets locally to meet the specific personalized requirements of different clients. Furthermore, to ensure fair benefit distribution, a model-aggregation method with optimized aggregation weights is used. This method reasonably allocates model-aggregation weights based on the contributions of clients, thereby achieving collaborative fairness. Experimental results demonstrate that, in three distinct data heterogeneity scenarios, employing personalized training through this framework exhibits higher accuracy compared to existing baseline methods. Simultaneously, the framework ensures collaborative fairness, fostering a more balanced and sustainable cooperation among IoT devices. |
| first_indexed | 2024-03-08T14:57:10Z |
| format | Article |
| id | doaj.art-428dba443e214b49b8975814dfec50e0 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-08T14:57:10Z |
| publishDate | 2023-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-428dba443e214b49b8975814dfec50e02024-01-10T15:08:31ZengMDPI AGSensors1424-82202023-12-012418810.3390/s24010088Personalized Fair Split Learning for Resource-Constrained Internet of ThingsHaitian Chen0Xuebin Chen1Lulu Peng2Yuntian Bai3College of Science, North China University of Science and Technology, Tangshan 063210, ChinaCollege of Science, North China University of Science and Technology, Tangshan 063210, ChinaCollege of Science, North China University of Science and Technology, Tangshan 063210, ChinaCollege of Science, North China University of Science and Technology, Tangshan 063210, ChinaWith the flourishing development of the Internet of Things (IoT), federated learning has garnered significant attention as a distributed learning method aimed at preserving the privacy of participant data. However, certain IoT devices, such as sensors, face challenges in effectively employing conventional federated learning approaches due to limited computational and storage resources, which hinder their ability to train complex local models. Additionally, in IoT environments, devices often face problems of data heterogeneity and uneven benefit distribution between them. To address these challenges, a personalized and fair split learning framework is proposed for resource-constrained clients. This framework first adopts a U-shaped structure, dividing the model to enable resource-constrained clients to offload subsets of the foundational model to a central server while retaining personalized model subsets locally to meet the specific personalized requirements of different clients. Furthermore, to ensure fair benefit distribution, a model-aggregation method with optimized aggregation weights is used. This method reasonably allocates model-aggregation weights based on the contributions of clients, thereby achieving collaborative fairness. Experimental results demonstrate that, in three distinct data heterogeneity scenarios, employing personalized training through this framework exhibits higher accuracy compared to existing baseline methods. Simultaneously, the framework ensures collaborative fairness, fostering a more balanced and sustainable cooperation among IoT devices.https://www.mdpi.com/1424-8220/24/1/88Internet of Thingsfederated learningsplit learningpersonalized modeldata heterogeneitycollaborative fairness |
| spellingShingle | Haitian Chen Xuebin Chen Lulu Peng Yuntian Bai Personalized Fair Split Learning for Resource-Constrained Internet of Things Sensors Internet of Things federated learning split learning personalized model data heterogeneity collaborative fairness |
| title | Personalized Fair Split Learning for Resource-Constrained Internet of Things |
| title_full | Personalized Fair Split Learning for Resource-Constrained Internet of Things |
| title_fullStr | Personalized Fair Split Learning for Resource-Constrained Internet of Things |
| title_full_unstemmed | Personalized Fair Split Learning for Resource-Constrained Internet of Things |
| title_short | Personalized Fair Split Learning for Resource-Constrained Internet of Things |
| title_sort | personalized fair split learning for resource constrained internet of things |
| topic | Internet of Things federated learning split learning personalized model data heterogeneity collaborative fairness |
| url | https://www.mdpi.com/1424-8220/24/1/88 |
| work_keys_str_mv | AT haitianchen personalizedfairsplitlearningforresourceconstrainedinternetofthings AT xuebinchen personalizedfairsplitlearningforresourceconstrainedinternetofthings AT lulupeng personalizedfairsplitlearningforresourceconstrainedinternetofthings AT yuntianbai personalizedfairsplitlearningforresourceconstrainedinternetofthings |