請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/101038完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 吳安宇 | zh_TW |
| dc.contributor.advisor | An-Yeu Wu | en |
| dc.contributor.author | 羅翊誠 | zh_TW |
| dc.contributor.author | Yi-Cheng Lo | en |
| dc.date.accessioned | 2025-11-26T16:34:07Z | - |
| dc.date.available | 2025-11-27 | - |
| dc.date.copyright | 2025-11-26 | - |
| dc.date.issued | 2025 | - |
| dc.date.submitted | 2025-09-04 | - |
| dc.identifier.citation | [1] C. Yang, S. Lan, L. Wang, W. Shen, and G. G. Q. Huang, “Big data driven edge-cloud collaboration architecture for cloud manufacturing: A software defined perspective,” IEEE Access, vol. 8, pp. 45938–45950, 2020.
[2] M. Armstrong, “Global data creation is about to explode.” https://www.statista.com/chart/17727/global-data-creation-forecasts/, 2019. [3] S. Kemp, “Digital 2023: Global overview report.” https://datareportal.com/reports/digital-2023-global-overview-report, 2023. [4] L. Dutta and S. Bharali, “TinyML meets IoT: A comprehensive survey,” Internet of Things, vol. 16, p. 100461, 2021. [5] M. AlSelek, J. M. Alcaraz-Calero, and Q. Wang, “Dynamic AI-IoT: Enabling updatable AI models in ultralow-power 5G IoT devices,” IEEE Internet of Things Journal, vol. 11, no. 8, pp. 14192–14205, 2024. [6] Y. Zhang, H. Yu, W. Zhou, and M. Man, “Application and research of IoT architecture for end-net-cloud edge computing,” Electronics, vol. 12, no. 1, 2023. [7] D. Donoho, “Compressed sensing,” IEEE Transactions on Information Theory, vol. 52, no. 4, pp. 1289–1306, 2006. [8] F. Chen, A. P. Chandrakasan, and V. M. Stojanovic, “Design and analysis of a hardware-efficient compressed sensing architecture for data compression in wireless sensors,” IEEE Journal of Solid-State Circuits, vol. 47, no. 3, pp. 744–756, 2012. [9] D. Gangopadhyay, E. G. Allstot, A. M. R. Dixon, K. Natarajan, S. Gupta, and D. J. Allstot, “Compressed sensing analog front-end for bio-sensor applications,” IEEE Journal of Solid-State Circuits, vol. 49, no. 2, pp. 426–438, 2014. [10] K. Xu, Y. Li, and F. Ren, “An energy-efficient compressive sensing framework incorporating online dictionary learning for long-term wireless health monitoring,” in 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 804–808, 2016. [11] R. Calderbank, S. Jafarpour, and R. Schapire, “Compressed learning: Universal sparse dimensionality reduction and learning in the measurement domain,” preprint, 2009. [12] H.-T. Li, C.-Y. Chou, Y.-T. Chen, S.-H. Wang, and A.-Y. Wu, “Robust and lightweight ensemble extreme learning machine engine based on eigenspace domain for compressed learning,” IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 66, no. 12, pp. 4699–4712, 2019. [13] W. Li, H. Chu, B. Huang, Y. Huan, L. Zheng, and Z. Zou, “Enabling on-device classification of ECG with compressed learning for health IoT,” Microelectronics Journal, vol. 115, p. 105188, 2021. [14] C.-Y. Chou, K.-C. Hsu, B.-H. Cho, K.-C. Chen, and A.-Y. A. Wu, “Low-complexity on-demand reconstruction for compressively sensed problematic signals,” IEEE Transactions on Signal Processing, vol. 68, pp. 4094–4107, 2020. [15] M. Zoni-Berisso, F. Lercari, T. Carazza, and S. Domenicucci, “Epidemiology of atrial fibrillation: European perspective,” Clinical epidemiology, pp. 213–220, 2014. [16] Y. Pati, R. Rezaiifar, and P. Krishnaprasad, “Orthogonal matching pursuit: recursive function approximation with applications to wavelet decomposition,” in Proceedings of 27th Asilomar Conference on Signals, Systems and Computers, pp. 40–44 vol.1, 1993. [17] D. Needell and J. A. Tropp, “CoSaMP: Iterative signal recovery from incomplete and inaccurate samples,” Applied and Computational Harmonic Analysis, vol. 26, no. 3, pp. 301–321, 2009. [18] A. Beck and M. Teboulle, “A fast iterative shrinkage-thresholding algorithm for linear inverse problems,” SIAM Journal on Imaging Sciences, vol. 2, no. 1, pp. 183–202, 2009. [19] D. L. Donoho, A. Maleki, and A. Montanari, “Message-passing algorithms for compressed sensing,” Proceedings of the National Academy of Sciences, vol. 106, no. 45, pp. 18914–18919, 2009. [20] U. Satija, B. Ramkumar, and M. S. Manikandan, “Automated ECG noise detection and classification system for unsupervised healthcare monitoring,” IEEE Journal of Biomedical and Health Informatics, vol. 22, no. 3, pp. 722–732, 2018. [21] K. Zhang, W. Zuo, Y. Chen, D. Meng, and L. Zhang, “Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising,” IEEE Transactions on Image Processing, vol. 26, no. 7, pp. 3142–3155, 2017. [22] S. Guo, Z. Yan, K. Zhang, W. Zuo, and L. Zhang, “Toward convolutional blind denoising of real photographs,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2019. [23] Y. Han, G. Huang, S. Song, L. Yang, H. Wang, and Y. Wang, “Dynamic neural networks: A survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 11, pp. 7436–7456, 2022. [24] D. Donoho, “De-noising by soft-thresholding,” IEEE Transactions on Information Theory, vol. 41, no. 3, pp. 613–627, 1995. [25] S. Poornachandra, “Wavelet-based denoising using subband dependent threshold for ECG signals,” Digital Signal Processing, vol. 18, no. 1, pp. 49–55, 2008. [26] Z. Chang, S. Liu, X. Xiong, Z. Cai, and G. Tu, “A survey of recent advances in edge-computing-powered artificial intelligence of things,” IEEE Internet of Things Journal, vol. 8, no. 18, pp. 13849–13875, 2021. [27] Y. Wang, C. Yang, S. Lan, L. Zhu, and Y. Zhang, “End-edge-cloud collaborative computing for deep learning: A comprehensive survey,” IEEE Communications Surveys & Tutorials, pp. 1–1, 2024. [28] Y. Kang, J. Hauswald, C. Gao, A. Rovinski, T. Mudge, J. Mars, and L. Tang, “Neurosurgeon: Collaborative intelligence between the cloud and mobile edge,” ACM SIGARCH Computer Architecture News, vol. 45, no. 1, pp. 615–629, 2017. [29] E. Li, L. Zeng, Z. Zhou, and X. Chen, “Edge AI: On-demand accelerating deep neural network inference via edge computing,” IEEE Transactions on Wireless Communications, vol. 19, no. 1, pp. 447–457, 2020. [30] M. Wołczyk, B. Wójcik, K. Bałazy, I. T. Podolak, J. Tabor, M. Śmieja, and T. Trzcinski, “Zero time waste: Recycling predictions in early exit neural networks,” Advances in Neural Information Processing Systems, vol. 34, pp. 2516–2528, 2021. [31] S. Teerapittayanon, B. McDanel, and H. Kung, “BranchyNet: Fast inference via early exiting from deep neural networks,” in 2016 23rd International Conference on Pattern Recognition (ICPR), pp. 2464–2469, 2016. [32] W. Shi, Y. Hou, S. Zhou, Z. Niu, Y. Zhang, and L. Geng, “Improving device-edge cooperative inference of deep learning via 2-step pruning,” in IEEE INFOCOM 2019 - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), pp. 1–6, 2019. [33] H. Zhou, W. Zhang, C. Wang, X. Ma, and H. Yu, “BBNet: a novel convolutional neural network structure in edge-cloud collaborative inference,” Sensors, vol. 21, no. 13, p. 4494, 2021. [34] Y. Chen, X. Wen, Y. Zhang, and W. Shi, “CCPrune: Collaborative channel pruning for learning compact convolutional networks,” Neurocomputing, vol. 451, pp. 35–45, 2021. [35] D. Hendrycks and K. Gimpel, “A baseline for detecting misclassified and out-of-distribution examples in neural networks,” CoRR, vol. abs/1610.02136, 2016. [36] C. Corbière, N. THOME, A. Bar-Hen, M. Cord, and P. Pérez, “Addressing failure prediction by learning model confidence,” in Advances in Neural Information Processing Systems, vol. 32, Curran Associates, Inc., 2019. [37] E. J. Candes and T. Tao, “Near-optimal signal recovery from random projections: Universal encoding strategies?,” IEEE Transactions on Information Theory, vol. 52, no. 12, pp. 5406–5425, 2006. [38] R. Durrant and A. Kabán, “Sharp generalization error bounds for randomly-projected classifiers,” in International Conference on Machine Learning, pp. 693–701, PMLR, 2013. [39] J. Mairal, F. Bach, J. Ponce, and G. Sapiro, “Online learning for matrix factorization and sparse coding.,” Journal of Machine Learning Research, vol. 11, no. 1, 2010. [40] J. Mairal, F. Bach, and J. Ponce, “Task-driven dictionary learning,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 34, no. 4, pp. 791–804, 2012. [41] K.-C. Hsu, B.-H. Cho, C.-Y. Chou, and A.-Y. A. Wu, “Low-complexity compressed analysis in eigenspace with limited labeled data for real-time electrocardiography telemonitoring,” in 2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP), pp. 459–463, IEEE, 2018. [42] K.-C. Chen, C.-Y. Chou, and A.-Y. Wu, “A tri-mode compressed analytics engine for low-power AF detection with on-demand EKG reconstruction,” IEEE Journal of Solid-State Circuits, vol. 56, no. 5, pp. 1608–1617, 2020. [43] Y. Abadade, A. Temouden, H. Bamoumen, N. Benamar, Y. Chtouki, and A. S. Hafid, “A comprehensive survey on TinyML,” IEEE Access, vol. 11, pp. 96892–96922, 2023. [44] J. Gawlikowski et al., “A survey of uncertainty in deep neural networks,” Artificial Intelligence Review, vol. 56, no. Suppl 1, pp. 1513–1589, 2023. [45] Y. He, T. Gan, W. Chen, and H. Wang, “Adaptive denoising by singular value decomposition,” IEEE Signal Processing Letters, vol. 18, no. 4, pp. 215–218, 2011. [46] W. Liu and W. Lin, “Additive white gaussian noise level estimation in SVD domain for images,” IEEE Transactions on Image processing, vol. 22, no. 3, pp. 872–883, 2012. [47] X. Li, S. Wang, and Y. Cai, “Tutorial: Complexity analysis of singular value decomposition and its variants,” arXiv preprint arXiv:1906.12085, 2019. [48] Y. Rachlin and D. Baron, “The secrecy of compressed sensing measurements,” in 2008 46th Annual Allerton Conference on Communication, Control, and Computing, pp. 813–817, 2008. [49] T. Bianchi, V. Bioglio, and E. Magli, “Analysis of one-time random projections for privacy preserving compressed sensing,” IEEE Transactions on Information Forensics and Security, vol. 11, no. 2, pp. 313–327, 2015. [50] C.-Y. Chou, E.-J. Chang, H.-T. Li, and A.-Y. Wu, “Low-complexity privacy-preserving compressive analysis using subspace-based dictionary for ECG telemonitoring system,” IEEE Transactions on Biomedical Circuits and Systems, vol. 12, no. 4, pp. 801–811, 2018. [51] K. Wang, Z. Liu, Y. Lin, J. Lin, and S. Han, “Haq: Hardware-aware automated quantization with mixed precision,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 8612–8620, 2019. [52] Y. Li and L. Guo, “Robust image fingerprinting via distortion-resistant sparse coding,” IEEE Signal Processing Letters, vol. 25, no. 1, pp. 140–144, 2018. [53] J. Yu, L. Yang, N. Xu, J. Yang, and T. Huang, “Slimmable neural networks,” arXiv preprint arXiv:1812.08928, 2018. [54] G. Moody, “A new method for detecting atrial fibrillation using RR intervals,” Computers in Cardiology, pp. 227–230, 1983. [55] R. Prasad, C. Dovrolis, M. Murray, and K. Claffy, “Bandwidth estimation: metrics, measurement techniques, and tools,” IEEE Network, vol. 17, no. 6, pp. 27–35, 2003. [56] Z. Hamici and W. Abu Elhaija, “Novel current unbalance estimation and diagnosis algorithms for condition monitoring with wireless sensor network and internet of things gateway,” IEEE Transactions on Industrial Informatics, vol. 15, no. 11, pp. 6080–6090, 2019. [57] K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” arXiv preprint arXiv:1409.1556, 2014. [58] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778, 2016. [59] A. Krizhevsky, G. Hinton, et al., “Learning multiple layers of features from tiny images,” 2009. [60] Y. Le and X. Yang, “Tiny imagenet visual recognition challenge,” CS 231N, vol. 7, no. 7, p. 3, 2015. [61] J. Yu and T. S. Huang, “Universally slimmable networks and improved training techniques,” in Proceedings of the IEEE/CVF international conference on computer vision, pp. 1803–1811, 2019. [62] X. Gao, Y. Zhao, Ł. Dudziak, R. Mullins, and C.-z. Xu, “Dynamic channel pruning: Feature boosting and suppression,” in International Conference on Learning Representations, 2018. [63] C. Li, G. Wang, B. Wang, X. Liang, Z. Li, and X. Chang, “Dynamic slimmable network,” in Proceedings of the IEEE/CVF Conference on computer vision and pattern recognition, pp. 8607–8617, 2021. [64] M. Kaya and H. Ş. Bilge, “Deep metric learning: A survey,” Symmetry, vol. 11, no. 9, 2019. [65] P. Khosla et al., “Supervised contrastive learning,” in Advances in Neural Information Processing Systems, vol. 33, pp. 18661–18673, 2020. [66] E. Hoffer and N. Ailon, “Deep metric learning using triplet network,” in Similarity-Based Pattern Recognition, pp. 84–92, Springer, 2015. | - |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/101038 | - |
| dc.description.abstract | 近年來，人工智慧 (AI) 技術快速的興起，強大的 AI 模型逐漸成為我們生活中不可或缺的一部份。人工智慧物聯網 (AIoT) 藉由結合邊緣運算和 AI 模型，在系統上達成傳感器-邊緣-雲的各層級協作，戰略性的將運算任務分配於不同層集中。這種階層式的系統，不僅能善用稀少的硬體資源，還能兼顧複雜問題的處理性，達成最佳的使用者體驗。然而，此議題上有許多潛在問題等待解決，包含在如今面臨的頻寬不足問題下，如何提高整體系統的運作效率和拓展性，是維持智慧物聯網生態系不可或缺的議題。
本論文就「IoT-邊緣端」和「邊緣-雲」兩項不同硬體等級的領域進行探討。此二領域同有著傳輸頻寬不足和運算資源限制的問題，而這也是過去文獻較為專注的部分。然而兩項研究還是有相異之處，在於前者有著 IoT 傳感器的訊號不穩定問題，而後者更多的是 AI模型-系統共同優化的問題。過去的文獻大多只專注於架構的設計，而缺乏特殊問題點的優化、整合性的分析、以及系統層面的優化。這會造成許多運算和網路資料傳輸資源的浪費，在 AIoT 系統內硬體資源非常珍貴的情況下，使的系統運作效率低落，無法完全的發揮系統協作的潛能。 本論文基於上述議題，分別對兩項研究領域進行問題整合，並提出嶄新的架構優化方法。對於IoT-邊緣端的雜訊干擾議題，而本論文提出一個動態調整的架構，先進行訊號品質評估，在進行最佳化的運算，進而使判斷變得更準確，從而提升硬體資源使用效率；本論文還提出一個輕量的訊號品質評估方法，不僅能精準的偵測雜訊大小，具有高度的可更新性，為系統的可擴展性加值。而對於邊緣-雲的共同優化問題，本論文提出一個兩階段的架構優化演算法，使的邊緣-雲中的 AI 模型能夠共同優化；此外，本論文更對網路資料的傳輸做更進一步的分析，使系統能更精確的傳輸真正關鍵的訊息，藉此極大化邊緣-雲的共同合作機制。基於這四項的深耕研究，本論文提及的架構和演算法打破過往的框架，有效的提升 AIoT 系統的效能。 | zh_TW |
| dc.description.abstract | In recent years, the advent of Artificial Intelligence (AI) has rapidly became a part of our daily lives. Artificial Intelligence of Things (AIoT) systems deftly allocate computing tasks across various layers, integrating edge computing with sophisticated AI models to facilitate seamless sensor-edge-cloud collaboration. This structured hierarchy not only optimizes limited hardware resources but also adeptly manages complex challenges to enhance user experiences. Nevertheless, numerous issues remain unresolved, including the enhancement of operational efficiency and scalability within the constraints of limited bandwidth, a critical factor for sustaining the AIoT ecosystem.
This dissertation examines the "IoT-edge" and "edge-cloud" domains. Both sectors confront common challenges such as insufficient bandwidth and restricted computational resources, themes prevalent in prior studies. However, distinctions emerge in the focus of these studies: IoT sensors frequently suffer from signal instability, whereas Edge-Cloud interactions primarily involve AI model-system co-optimization. Previous research has concentrated on architectural design yet often overlooks the optimization system-level enhancements. This oversight leads to inefficient utilization of computing capabilities and network data transfers, thereby undermining system efficacy and failing to leverage the full potential of collaborative operations in resource-constrained AIoT environments. To address the deficiencies, this dissertation synthesizes issues from both research domains and introduces an innovative architectural optimization strategy. First, to counteract IoT-edge noise interference, we propose a dynamic architecture that initiates with signal quality assessment to select the best model. We also introduce a lightweight signal assessment technique that not only accurately gauges noise levels but is highly adaptable. Second, regarding the edge-cloud co-optimization challenge, we advocate for a two-stage architectural optimization algorithm that facilitates the joint optimization of AI models among edge and cloud. In addition, we further delve into data transmission optimizations, enabling the system to more effectively relay critical information and maximize the cooperative mechanisms between edge and cloud components. Through these comprehensive studies, the proposed architectures and algorithms transcend traditional frameworks, significantly improving the performance of AIoT systems. | en |
| dc.description.provenance | Submitted by admin ntu (admin@lib.ntu.edu.tw) on 2025-11-26T16:34:07Z No. of bitstreams: 0 | en |
| dc.description.provenance | Made available in DSpace on 2025-11-26T16:34:07Z (GMT). No. of bitstreams: 0 | en |
| dc.description.tableofcontents | 致謝 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i
摘要 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv Chapter 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1.1 The AIoT: From Cloud-Centric to Hierarchical Collaboration . . . . . . . . . . . . . . 1 1.1.2 Hierarchical Computing Paradigm and Resource Limitations . . . . . . . . . . . . . . . 3 1.2 Efficient IoT-Edge System by Compressed Analysis . . . . . . . . . . . . . . . . . . . . 4 1.2.1 Compressed Sensing for Joint Sampling and Compression . . . . . . . . . . . . . . . . . 4 1.2.2 Compressed Analysis for Early Filtering . . . . . . . . . . . . . . . . . . . . . . . . 6 1.2.3 Design Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.2.4 Research Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.3 Collaborative Neural Networks in the Edge-Cloud . . . . . . . . . . . . . . . . . . . . . 11 1.3.1 Distributed CNN Computing by Edge-Cloud Collaboration . . . . . . . . . . . . . . . . . 11 1.3.2 Bandwidth Reduction by Early-Exiting Architecture . . . . . . . . . . . . . . . . . . . 12 1.3.3 Design Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.3.4 Research Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 1.4 Dissertation Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Chapter 2 Review of Related Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.1 Preliminary of CA for ECG Signal Monitoring . . . . . . . . . . . . . . . . . . . . . . . 19 2.1.1 Terminology of Compressed Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.1.2 Architecture of ECG Monitoring by CA . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.2 Related Works of CA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.2.1 Development of CA Architectures . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.2.2 Related Works of the Hardware-Efficient CA Architecture . . . . . . . . . . . . . . . . 22 2.3 Preliminary of Edge-Cloud Collaboration on CNN . . . . . . . . . . . . . . . . . . . . . 23 2.4 Related Works of Edge-Cloud Collaborative System with On-Demand Offloading Mechanism . . 25 2.4.1 Related Works of Pruning on the Edge Model . . . . . . . . . . . . . . . . . . . . . . 25 2.4.2 Related Works of the Offloading Policy . . . . . . . . . . . . . . . . . . . . . . . . 26 2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Chapter 3 Efficient Noise-Level Estimation on the Compressed Domain . . . . . . . . . . . . . 29 3.1 Proposed SVD-Based Noise-Level Estimation . . . . . . . . . . . . . . . . . . . . . . . . 29 3.1.1 Challenges of Noise-Level Estimation on Compressed Domain . . . . . . . . . . . . . . . 29 3.1.2 Noise-Level Estimation on the Compressed Data . . . . . . . . . . . . . . . . . . . . . 30 3.1.3 Deviation Correction by Removal of Residual Signal . . . . . . . . . . . . . . . . . . 32 3.2 Efficient Refreshing Mechanism for Noise Estimator by Transfer Learning . . . . . . . . . 34 3.2.1 Challenges of Noise-Level Estimation on Compressed Domain . . . . . . . . . . . . . . . 34 3.2.2 Privacy Issue and the Compression Matrix Refreshing . . . . . . . . . . . . . . . . . . 35 3.3 Performance Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 3.3.1 Experimental Setups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 3.3.2 Experimental Results on the Accuracy of the Proposed SVD-NE . . . . . . . . . . . . . . 37 3.3.3 Experimental Results of the Proposed Refreshing Mechanism . . . . . . . . . . . . . . . 39 3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Chapter 4 Noise-Level Aware Compressed Analysis for Robust Prediction . . . . . . . . . . . . 41 4.1 Observation of Noise Effect on CA Models . . . . . . . . . . . . . . . . . . . . . . . . 41 4.1.1 Properties of the Features in the TS-CAR framework . . . . . . . . . . . . . . . . . . 42 4.1.2 Accuracy of TS-CAR Models with Different Sizes . . . . . . . . . . . . . . . . . . . . 45 4.2 Proposed Noise-Level Aware Compressed Analysis . . . . . . . . . . . . . . . . . . . . . 47 4.2.1 Training of the Proposed NA-CA . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 4.2.2 Inference of the Proposed NA-CA . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 4.3 Light-Weight Model Design for the IoT’s Constraints . . . . . . . . . . . . . . . . . . . 52 4.4 Performance Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 4.4.1 Experimental Setups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 4.4.2 Experimental Results on the Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . 56 4.4.3 Hardware Evaluation of the Proposed Architecture . . . . . . . . . . . . . . . . . . . 57 4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Chapter 5 CNN with the Efficient Edge-Cloud Collaborative Computing . . . . . . . . . . . . . 61 5.1 Rethink of the Edge-Cloud Collaboration of CNNs with On-Demand Offloading Mechanism . . . 61 5.2 Proposed Efficient Edge-Cloud Collaborative System . . . . . . . . . . . . . . . . . . . 62 5.2.1 Architecture Breakdown of the Proposed EECCS . . . . . . . . . . . . . . . . . . . . . 62 5.2.2 Training of the Proposed EECCS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 5.2.3 Inference of the Proposed EECCS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 5.3 Constraints-Aware Trainable Pruning . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 5.3.1 Automatic Pruning Based on Trainable Gates . . . . . . . . . . . . . . . . . . . . . . 67 5.3.2 System-Level Optimization by Performance Predictor . . . . . . . . . . . . . . . . . . 70 5.4 Performance Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 5.4.1 Experimental Setups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 5.4.2 Experimental Results on the Accuracy of the Proposed Architecture . . . . . . . . . . . 72 5.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Chapter 6 Design of an Efficient Controller with System-Level Optimization . . . . . . . . . 77 6.1 Towards Leveraging the On-Demand Offloading . . . . . . . . . . . . . . . . . . . . . . . 77 6.2 Dynamic Inference for Acute Offloading Information . . . . . . . . . . . . . . . . . . . 78 6.2.1 Rethink of the Offloading Information from Edge and Cloud . . . . . . . . . . . . . . . 78 6.2.2 Enabling Static Pruning and Dynamic Pruning in EECCS . . . . . . . . . . . . . . . . . 80 6.2.3 Training of the Information Controller . . . . . . . . . . . . . . . . . . . . . . . . 81 6.2.4 Inference of the Information Controller . . . . . . . . . . . . . . . . . . . . . . . . 83 6.3 Trainable Policy to Optimize Offloading Decision . . . . . . . . . . . . . . . . . . . . 85 6.3.1 Rethink of the Offloading Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 6.3.2 Reformulation of Criteria for System-Level Optimization . . . . . . . . . . . . . . . . 87 6.3.3 Advanceing Embedding Features to Enhance Representation . . . . . . . . . . . . . . . . 89 6.4 Performance Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 6.4.1 Experimental Setups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 6.4.2 Experimental Results of the Dynamic Inference Module . . . . . . . . . . . . . . . . . 92 6.4.3 Experimental Results of Trainable Policy . . . . . . . . . . . . . . . . . . . . . . . 95 6.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Chapter 7 Conclusions and Future Works . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 7.1 Main Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 7.2 Future Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 | - |
| dc.language.iso | zh_TW | - |
| dc.subject | 智慧物聯網 | - |
| dc.subject | 階層式分析 | - |
| dc.subject | 壓縮域分析 | - |
| dc.subject | 模型壓縮 | - |
| dc.subject | 系統優化 | - |
| dc.subject | Artificial Intelligence of Things | - |
| dc.subject | hierarchical analysis | - |
| dc.subject | compressed analysis | - |
| dc.subject | model compression | - |
| dc.subject | system-level optimization | - |
| dc.title | 適用於智慧物聯網應用之高效能邊緣雲共同計算技術 | zh_TW |
| dc.title | Efficient Edge-Cloud Collaborative Computing for Intelligent IoT Applications | en |
| dc.type | Thesis | - |
| dc.date.schoolyear | 114-1 | - |
| dc.description.degree | 博士 | - |
| dc.contributor.oralexamcommittee | 盧奕璋;楊家驤;蔡佩芸;賴以威;沈中安;黃元豪 | zh_TW |
| dc.contributor.oralexamcommittee | Yi-Chang Lu;Chia-Hsiang Yang;Pei-Yun Tsai;I-Wei Lai;Chung-An Shen;Yuan-Hao Huang | en |
| dc.subject.keyword | 智慧物聯網,階層式分析壓縮域分析模型壓縮系統優化 | zh_TW |
| dc.subject.keyword | Artificial Intelligence of Things,hierarchical analysiscompressed analysismodel compressionsystem-level optimization | en |
| dc.relation.page | 111 | - |
| dc.identifier.doi | 10.6342/NTU202503524 | - |
| dc.rights.note | 同意授權(限校園內公開) | - |
| dc.date.accepted | 2025-09-05 | - |
| dc.contributor.author-college | 電機資訊學院 | - |
| dc.contributor.author-dept | 電子工程學研究所 | - |
| dc.date.embargo-lift | 2025-10-01 | - |
| 顯示於系所單位： | 電子工程學研究所 | |
文件中的檔案：
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| ntu-114-1.pdf 授權僅限NTU校內IP使用（校園外請利用VPN校外連線服務） | 7.37 MB | Adobe PDF |
系統中的文件，除了特別指名其著作權條款之外，均受到著作權保護，並且保留所有的權利。