NEO: Aprendizado Federado com Meta-Aprendizado Dinâmico para Redes Veiculares
Resumo
As Redes Veiculares Ad Hoc (VANETs) permitem a comunicação direta entre veículos e a infraestrutura, contribuindo para a segurança e a eficiência do tráfego. No entanto, desafios como a alta mobilidade e as restrições computacionais dificultam a aplicação de técnicas de Aprendizado de Máquina (ML) nesse contexto. O Aprendizado Federado (FL) possibilita o treinamento distribuído de modelos preservando a privacidade dos dados, mas frequentemente exige um grande número de rodadas de comunicação para atingir uma boa convergência. Nesse sentido, este trabalho propõe uma abordagem adaptativa de Meta-Aprendizado Federado, baseada em Redes Convolucionais de Grafos (GCNs), para a previsão de tendências de velocidade em VANETs. A solução ajusta dinamicamente a intensidade do Meta-Aprendizado, otimizando a eficiência do treinamento e reduzindo a sobrecarga computacional. Resultados experimentais indicam que a abordagem reduz o erro preditivo em 5% e diminui o tempo de treinamento em 8%, quando comparada a outras abordagens do estado da arte, demonstrando-se uma alternativa viável para aplicações em VANETs com restrições de recursos.
Palavras-chave:
Aprendizado Federado para Computação Urbana, Cidades Inteligentes, Internet das Coisas, Meta-Aprendizado, Planejamento Urbano Usando Big Data, Sistemas de Transporte Inteligentes
Referências
Al-shareeda, M. A., Alazzawi, M. A., Anbar, M., Manickam, S., and Al-Ani, A. K.(2021). A comprehensive survey on vehicular ad hoc networks (vanets). In 2021 International Conference on Advanced Computer Applications (ACA), pages 156–160.
California Department of Transportation (Caltrans) (2025). PeMS data source. [link]. Acesso em: 31 mar. 2025.
Chellapandi, V. P., Yuan, L., Brinton, C. G., Żak, S. H., and Wang, Z. (2024). Federated learning for connected and automated vehicles: A survey of existing approaches and challenges. IEEE Transactions on Intelligent Vehicles, 9(1):119–137.
Chen, F., Luo, M., Dong, Z., Li, Z., and He, X. (2019). Federated meta-learning with fast convergence and efficient communication.
de Souza, A. M., Braun, T., Botega, L. C., Villas, L. A., and Loureiro, A. A. F. (2020). Safe and sound: Driver safety-aware vehicle re-routing based on spatiotemporal information. IEEE Transactions on Intelligent Transportation Systems, 21(9):3973–3989.
de Souza, A. M., Maciel, F., da Costa, J. B., Bittencourt, L. F., Cerqueira, E., Loureiro, A. A., and Villas, L. A. (2024). Adaptive client selection with personalization for communication efficient federated learning. Ad Hoc Networks, 157:103462.
Fei, X. and Ling, Q. (2023). Attention-based global and local spatial-temporal graph convolutional network for vehicle emission prediction. Neurocomputing, 521:41–55.
Feng, X., Sun, H., Liu, S., Guo, J., and Zheng, H. (2024). Federated meta-learning on graph for traffic flow prediction. IEEE Transactions on Vehicular Technology, 73(12):19526–19538.
Finn, C., Abbeel, P., and Levine, S. (2017). Model-agnostic meta-learning for fast adaptation of deep networks.
He, Y., Wang, Y., Lin, Q., and Li, J. (2022). Meta-hierarchical reinforcement learning (mhrl)-based dynamic resource allocation for dynamic vehicular networks. IEEE Transactions on Vehicular Technology, 71(4):3495–3506.
Lee, R., Kim, M., Li, D., Qiu, X., Hospedales, T., Huszár, F., and Lane, N. D. (2023). Fedl2p: Federated learning to personalize.
Li, N., Zhao, S., Feng, Y., and Han, F. (2024). Mgsta: Meta learning based graph convolutional stacked temporal attention neural network for traffic flow forecasting. In 2024 International Joint Conference on Neural Networks (IJCNN), pages 1–8.
Li, W. and Wang, S. (2022). Federated meta-learning for spatial-temporal prediction. Neural Computing and Applications, 34(13):10355–10374.
Nayomi, B. D. and Jyothsna, V. (2024). A comprehensive survey on deep learning for enhanced node position prediction in vehicular ad-hoc networks. In 2024 4th International Conference on Ubiquitous Computing and Intelligent Information Systems (ICUIS), pages 602–609.
Nichol, A., Achiam, J., and Schulman, J. (2018). On first-order meta-learning algorithms.
Qi, T., Chen, L., Li, G., Li, Y., and Wang, C. (2023). Fedagcn: A traffic flow prediction framework based on federated learning and asynchronous graph convolutional network. Applied Soft Computing, 138:110175.
Rajeswaran, A., Finn, C., Kakade, S., and Levine, S. (2019). Meta-learning with implicit gradients.
Souza, A., Bittencourt, L., Cerqueira, E., Loureiro, A., and Villas, L. (2023). Dispositivos, eu escolho vocês: Seleção de clientes adaptativa para comunicação eficiente em aprendizado federado. In Anais do XLI Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos, pages 1–14, Porto Alegre, RS, Brasil. SBC.
Tan, K., Bremner, D., Le Kernec, J., Zhang, L., and Imran, M. (2022). Machine learning in vehicular networking: An overview. Digital Communications and Networks, 8(1):18–24.
Valente, R., Senna, C., Rito, P., and Sargento, S. (2023). Embedded federated learning for vanet environments. Applied Sciences, 13(4).
Wang, H., Zhang, R., Cheng, X., and Yang, L. (2022). Federated spatio-temporal traffic flow prediction based on graph convolutional network. In 2022 14th International Conference on Wireless Communications and Signal Processing (WCSP), pages 221–225.
Wen, J., Zhang, Z., Lan, Y., Cui, Z., Cai, J., and Zhang, W. (2023). A survey on federated learning: challenges and applications. International Journal of Machine Learning and Cybernetics, 14(2):513–535.
Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., and Yu, P. S. (2021). A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems, 32(1):4–24.
You, L., Chen, Q., Qu, H., Zhu, R., Yan, J., Santi, P., and Ratti, C. (2024). Fmgcn: Federated meta learning-augmented graph convolutional network for ev charging demand forecasting. IEEE Internet of Things Journal, 11(14):24452–24466.
Yu, B., Yin, H., and Zhu, Z. (2018). Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting. In Proceedings of the 27th International Joint Conference on Artificial Intelligence (IJCAI).
California Department of Transportation (Caltrans) (2025). PeMS data source. [link]. Acesso em: 31 mar. 2025.
Chellapandi, V. P., Yuan, L., Brinton, C. G., Żak, S. H., and Wang, Z. (2024). Federated learning for connected and automated vehicles: A survey of existing approaches and challenges. IEEE Transactions on Intelligent Vehicles, 9(1):119–137.
Chen, F., Luo, M., Dong, Z., Li, Z., and He, X. (2019). Federated meta-learning with fast convergence and efficient communication.
de Souza, A. M., Braun, T., Botega, L. C., Villas, L. A., and Loureiro, A. A. F. (2020). Safe and sound: Driver safety-aware vehicle re-routing based on spatiotemporal information. IEEE Transactions on Intelligent Transportation Systems, 21(9):3973–3989.
de Souza, A. M., Maciel, F., da Costa, J. B., Bittencourt, L. F., Cerqueira, E., Loureiro, A. A., and Villas, L. A. (2024). Adaptive client selection with personalization for communication efficient federated learning. Ad Hoc Networks, 157:103462.
Fei, X. and Ling, Q. (2023). Attention-based global and local spatial-temporal graph convolutional network for vehicle emission prediction. Neurocomputing, 521:41–55.
Feng, X., Sun, H., Liu, S., Guo, J., and Zheng, H. (2024). Federated meta-learning on graph for traffic flow prediction. IEEE Transactions on Vehicular Technology, 73(12):19526–19538.
Finn, C., Abbeel, P., and Levine, S. (2017). Model-agnostic meta-learning for fast adaptation of deep networks.
He, Y., Wang, Y., Lin, Q., and Li, J. (2022). Meta-hierarchical reinforcement learning (mhrl)-based dynamic resource allocation for dynamic vehicular networks. IEEE Transactions on Vehicular Technology, 71(4):3495–3506.
Lee, R., Kim, M., Li, D., Qiu, X., Hospedales, T., Huszár, F., and Lane, N. D. (2023). Fedl2p: Federated learning to personalize.
Li, N., Zhao, S., Feng, Y., and Han, F. (2024). Mgsta: Meta learning based graph convolutional stacked temporal attention neural network for traffic flow forecasting. In 2024 International Joint Conference on Neural Networks (IJCNN), pages 1–8.
Li, W. and Wang, S. (2022). Federated meta-learning for spatial-temporal prediction. Neural Computing and Applications, 34(13):10355–10374.
Nayomi, B. D. and Jyothsna, V. (2024). A comprehensive survey on deep learning for enhanced node position prediction in vehicular ad-hoc networks. In 2024 4th International Conference on Ubiquitous Computing and Intelligent Information Systems (ICUIS), pages 602–609.
Nichol, A., Achiam, J., and Schulman, J. (2018). On first-order meta-learning algorithms.
Qi, T., Chen, L., Li, G., Li, Y., and Wang, C. (2023). Fedagcn: A traffic flow prediction framework based on federated learning and asynchronous graph convolutional network. Applied Soft Computing, 138:110175.
Rajeswaran, A., Finn, C., Kakade, S., and Levine, S. (2019). Meta-learning with implicit gradients.
Souza, A., Bittencourt, L., Cerqueira, E., Loureiro, A., and Villas, L. (2023). Dispositivos, eu escolho vocês: Seleção de clientes adaptativa para comunicação eficiente em aprendizado federado. In Anais do XLI Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos, pages 1–14, Porto Alegre, RS, Brasil. SBC.
Tan, K., Bremner, D., Le Kernec, J., Zhang, L., and Imran, M. (2022). Machine learning in vehicular networking: An overview. Digital Communications and Networks, 8(1):18–24.
Valente, R., Senna, C., Rito, P., and Sargento, S. (2023). Embedded federated learning for vanet environments. Applied Sciences, 13(4).
Wang, H., Zhang, R., Cheng, X., and Yang, L. (2022). Federated spatio-temporal traffic flow prediction based on graph convolutional network. In 2022 14th International Conference on Wireless Communications and Signal Processing (WCSP), pages 221–225.
Wen, J., Zhang, Z., Lan, Y., Cui, Z., Cai, J., and Zhang, W. (2023). A survey on federated learning: challenges and applications. International Journal of Machine Learning and Cybernetics, 14(2):513–535.
Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., and Yu, P. S. (2021). A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems, 32(1):4–24.
You, L., Chen, Q., Qu, H., Zhu, R., Yan, J., Santi, P., and Ratti, C. (2024). Fmgcn: Federated meta learning-augmented graph convolutional network for ev charging demand forecasting. IEEE Internet of Things Journal, 11(14):24452–24466.
Yu, B., Yin, H., and Zhu, Z. (2018). Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting. In Proceedings of the 27th International Joint Conference on Artificial Intelligence (IJCAI).
Publicado
19/05/2025
Como Citar
BRAUN, Carnot; JARCZEWSKI, Rafael O.; DA COSTA, Joahannes B. D.; DE SOUZA, Allan M..
NEO: Aprendizado Federado com Meta-Aprendizado Dinâmico para Redes Veiculares. In: WORKSHOP DE COMPUTAÇÃO URBANA (COURB), 9. , 2025, Natal/RN.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2025
.
p. 293-306.
ISSN 2595-2706.
DOI: https://doi.org/10.5753/courb.2025.9588.
