能源採集智慧物聯網設備提供系統永續性，並克服不穩定的傳輸挑戰。然而，在能源採集設備上執行多個深度神經網絡推理引來排程與資源管理的挑戰。為了保持微控制器和加速器之間的數據一致性，推理任務通常是不可搶占的，進而延後其他應用程序的執行。此外，耗費能源的推理任務在能源採集系統上的斷續執行行為大幅降低推理任務的響應能力。

我們提出了一個排程器結合半搶占式執行與提早離開的政策以權衡推理任務的時效性與正確性，並提出一個動態能源管理機制結合排程控制器，用以管理推理任務以及周邊任務的斷續執行行為。我們的方法實現在真實平台上，並與相關研究工作進行比較，其評估結果顯示我們的方法顯著地提高工作時效內完成率和工作完成率。

Energy-harvesting intelligent IoT devices are getting attracted to provide sustainable development and overcome unstable communication. However, executing the multiple latency-constraints deep neural network (DNN) inferences suffers both scheduling and energy management challenges. To maintain data consistency between MCU and accelerator, the non-preemptible inference task execution results in a longer blocking time for other applications. The energy-hungry inference task suffers significant power failure on energy-harvesting devices to meet the timely response. We propose a scheduler combining the preemption point and early exit to trade off the accuracy and responsiveness of the inferences with the admission control. A dynamic energy isolation concept is then presented to cooperate with the priority-driven scheduler for managing the intermittent executions on the harvesting devices with peripherals. Finally, we implement our approach on the real platform and compare the state-of-art work, and our proposed approach significantly increases the performance of the meet ratio and progress ratios.

[1] T. Wan, Y. Karimi, M. Stanacevi ´ c, and E. Salman, “Ac computing methodology for ´
rf-powered iot devices,” IEEE Transactions on Very Large Scale Integration (VLSI)
Systems, vol. 27, no. 5, pp. 1017–1028, 2019.
[2] Y. Tan, Y. Shiiki, and H. Ishikuro, “Optimization of gate voltage in capacitive dc–dc
converters for thermoelectric energy harvesting,” IEEE Transactions on Very Large
Scale Integration (VLSI) Systems, vol. 30, no. 4, pp. 463–473, 2022.
[3] M. Rasheduzzaman, P. B. Pillai, A. N. C. Mendoza, and M. M. De Souza, “A study
of the performance of solar cells for indoor autonomous wireless sensors,” in 2016
10th International Symposium on Communication Systems, Networks and Digital Signal Processing (CSNDSP), pp. 1–6, 2016.
[4] N. A. Bhatti, M. H. Alizai, A. A. Syed, and L. Mottola, “Energy harvesting and wireless
transfer in sensor network applications: Concepts and experiences,” ACM Trans. Sen.
Netw., vol. 12, aug 2016.
[5] M. Ebrahimi, M. Khoshtaghaza, S. Minaei, and B. Jamshidi, “Vision-based pest detection based on svm classification method,” Computers and Electronics in Agriculture,
vol. 137, pp. 52–58, 2017.
[6] G. Gobieski, B. Lucia, and N. Beckmann, “Intelligence beyond the edge: Inference
on intermittent embedded systems,” in Proceedings of the Twenty-Fourth International
Conference on Architectural Support for Programming Languages and Operating Systems, pp. 199–213, 2019.
[7] Y. Wu, Z. Wang, Z. Jia, Y. Shi, and J. Hu, “Intermittent inference with nonuniformly
compressed multi-exit neural network for energy harvesting powered devices,” in 2020
57th ACM/IEEE Design Automation Conference (DAC), pp. 1–6, 2020.
[8] C.-K. Kang, H. R. Mendis, C.-H. Lin, M.-S. Chen, and P.-C. Hsiu, “Everything leaves
footprints: Hardware accelerated intermittent deep inference,” IEEE Transactions on
Computer-Aided Design of Integrated Circuits and Systems, vol. 39, no. 11, pp. 3479–
3491, 2020.
[9] C.-K. Kang, H. R. Mendis, C.-H. Lin, M.-S. Chen, and P.-C. Hsiu, “More is less: Model
augmentation for intermittent deep inference,” ACM Trans. Embed. Comput. Syst., dec
2021. Just Accepted.
[10] M. B. Hoy, “Alexa, siri, cortana, and more: An introduction to voice assistants,” Medical Reference Services Quarterly, vol. 37, no. 1, pp. 81–88, 2018. PMID: 29327988.
[11] D. Gutierrez-Galan, J. P. Dominguez-Morales, E. Cerezuela-Escudero, A. RiosNavarro, R. Tapiador-Morales, M. Rivas-Perez, M. Dominguez-Morales, A. JimenezFernandez, and A. Linares-Barranco, “Embedded neural network for real-time animal
behavior classification,” Neurocomputing, vol. 272, pp. 17–26, 2018.
[12] R. Srinivasan, M. T. Islam, B. Islam, Z. Wang, T. Sookoor, O. Gnawali, and S. Nirjon,
“Preventive maintenance of centralized hvac systems: use of acoustic sensors, feature
extraction, and unsupervised learning,” in Proceedings of the 15th IBPSA Conference,
pp. 2518–2524, 2017.
[13] B. Islam and S. Nirjon, “Zygarde: Time-sensitive on-device deep inference and adaptation on intermittently-powered systems,” Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., vol. 4, Sept. 2020.
[14] W.-M. Chen, T.-W. Kuo, and P.-C. Hsiu, “Enabling failure-resilient intermittent systems without runtime checkpointing,” IEEE Transactions on Computer-Aided Design
of Integrated Circuits and Systems, vol. 39, no. 12, pp. 4399–4412, 2020.
[7] Y. Wu, Z. Wang, Z. Jia, Y. Shi, and J. Hu, “Intermittent inference with nonuniformly
compressed multi-exit neural network for energy harvesting powered devices,” in 2020
57th ACM/IEEE Design Automation Conference (DAC), pp. 1–6, 2020.
[8] C.-K. Kang, H. R. Mendis, C.-H. Lin, M.-S. Chen, and P.-C. Hsiu, “Everything leaves
footprints: Hardware accelerated intermittent deep inference,” IEEE Transactions on
Computer-Aided Design of Integrated Circuits and Systems, vol. 39, no. 11, pp. 3479–
3491, 2020.
[9] C.-K. Kang, H. R. Mendis, C.-H. Lin, M.-S. Chen, and P.-C. Hsiu, “More is less: Model
augmentation for intermittent deep inference,” ACM Trans. Embed. Comput. Syst., dec
2021. Just Accepted.
[10] M. B. Hoy, “Alexa, siri, cortana, and more: An introduction to voice assistants,” Medical Reference Services Quarterly, vol. 37, no. 1, pp. 81–88, 2018. PMID: 29327988.
[11] D. Gutierrez-Galan, J. P. Dominguez-Morales, E. Cerezuela-Escudero, A. RiosNavarro, R. Tapiador-Morales, M. Rivas-Perez, M. Dominguez-Morales, A. JimenezFernandez, and A. Linares-Barranco, “Embedded neural network for real-time animal
behavior classification,” Neurocomputing, vol. 272, pp. 17–26, 2018.
[12] R. Srinivasan, M. T. Islam, B. Islam, Z. Wang, T. Sookoor, O. Gnawali, and S. Nirjon,
“Preventive maintenance of centralized hvac systems: use of acoustic sensors, feature
extraction, and unsupervised learning,” in Proceedings of the 15th IBPSA Conference,
pp. 2518–2524, 2017.
[13] B. Islam and S. Nirjon, “Zygarde: Time-sensitive on-device deep inference and adaptation on intermittently-powered systems,” Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., vol. 4, Sept. 2020.
[14] W.-M. Chen, T.-W. Kuo, and P.-C. Hsiu, “Enabling failure-resilient intermittent systems without runtime checkpointing,” IEEE Transactions on Computer-Aided Design
of Integrated Circuits and Systems, vol. 39, no. 12, pp. 4399–4412, 2020.
[7] Y. Wu, Z. Wang, Z. Jia, Y. Shi, and J. Hu, “Intermittent inference with nonuniformly
compressed multi-exit neural network for energy harvesting powered devices,” in 2020
57th ACM/IEEE Design Automation Conference (DAC), pp. 1–6, 2020.
[8] C.-K. Kang, H. R. Mendis, C.-H. Lin, M.-S. Chen, and P.-C. Hsiu, “Everything leaves
footprints: Hardware accelerated intermittent deep inference,” IEEE Transactions on
Computer-Aided Design of Integrated Circuits and Systems, vol. 39, no. 11, pp. 3479–
3491, 2020.
[9] C.-K. Kang, H. R. Mendis, C.-H. Lin, M.-S. Chen, and P.-C. Hsiu, “More is less: Model
augmentation for intermittent deep inference,” ACM Trans. Embed. Comput. Syst., dec
2021. Just Accepted.
[10] M. B. Hoy, “Alexa, siri, cortana, and more: An introduction to voice assistants,” Medical Reference Services Quarterly, vol. 37, no. 1, pp. 81–88, 2018. PMID: 29327988.
[11] D. Gutierrez-Galan, J. P. Dominguez-Morales, E. Cerezuela-Escudero, A. RiosNavarro, R. Tapiador-Morales, M. Rivas-Perez, M. Dominguez-Morales, A. JimenezFernandez, and A. Linares-Barranco, “Embedded neural network for real-time animal
behavior classification,” Neurocomputing, vol. 272, pp. 17–26, 2018.
[12] R. Srinivasan, M. T. Islam, B. Islam, Z. Wang, T. Sookoor, O. Gnawali, and S. Nirjon,
“Preventive maintenance of centralized hvac systems: use of acoustic sensors, feature
extraction, and unsupervised learning,” in Proceedings of the 15th IBPSA Conference,
pp. 2518–2524, 2017.
[13] B. Islam and S. Nirjon, “Zygarde: Time-sensitive on-device deep inference and adaptation on intermittently-powered systems,” Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., vol. 4, Sept. 2020.
[14] W.-M. Chen, T.-W. Kuo, and P.-C. Hsiu, “Enabling failure-resilient intermittent systems without runtime checkpointing,” IEEE Transactions on Computer-Aided Design
of Integrated Circuits and Systems, vol. 39, no. 12, pp. 4399–4412, 2020.
[7] Y. Wu, Z. Wang, Z. Jia, Y. Shi, and J. Hu, “Intermittent inference with nonuniformly
compressed multi-exit neural network for energy harvesting powered devices,” in 2020
57th ACM/IEEE Design Automation Conference (DAC), pp. 1–6, 2020.
[8] C.-K. Kang, H. R. Mendis, C.-H. Lin, M.-S. Chen, and P.-C. Hsiu, “Everything leaves
footprints: Hardware accelerated intermittent deep inference,” IEEE Transactions on
Computer-Aided Design of Integrated Circuits and Systems, vol. 39, no. 11, pp. 3479–
3491, 2020.
[9] C.-K. Kang, H. R. Mendis, C.-H. Lin, M.-S. Chen, and P.-C. Hsiu, “More is less: Model
augmentation for intermittent deep inference,” ACM Trans. Embed. Comput. Syst., dec
2021. Just Accepted.
[10] M. B. Hoy, “Alexa, siri, cortana, and more: An introduction to voice assistants,” Medical Reference Services Quarterly, vol. 37, no. 1, pp. 81–88, 2018. PMID: 29327988.
[11] D. Gutierrez-Galan, J. P. Dominguez-Morales, E. Cerezuela-Escudero, A. RiosNavarro, R. Tapiador-Morales, M. Rivas-Perez, M. Dominguez-Morales, A. JimenezFernandez, and A. Linares-Barranco, “Embedded neural network for real-time animal
behavior classification,” Neurocomputing, vol. 272, pp. 17–26, 2018.
[12] R. Srinivasan, M. T. Islam, B. Islam, Z. Wang, T. Sookoor, O. Gnawali, and S. Nirjon,
“Preventive maintenance of centralized hvac systems: use of acoustic sensors, feature
extraction, and unsupervised learning,” in Proceedings of the 15th IBPSA Conference,
pp. 2518–2524, 2017.
[13] B. Islam and S. Nirjon, “Zygarde: Time-sensitive on-device deep inference and adaptation on intermittently-powered systems,” Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., vol. 4, Sept. 2020.
[14] W.-M. Chen, T.-W. Kuo, and P.-C. Hsiu, “Enabling failure-resilient intermittent systems without runtime checkpointing,” IEEE Transactions on Computer-Aided Design
of Integrated Circuits and Systems, vol. 39, no. 12, pp. 4399–4412, 2020.
[39] Y.-C. Lin, P.-C. Hsiu, and T.-W. Kuo, “Autonomous i/o for intermittent iot systems,”
in 2019 IEEE/ACM International Symposium on Low Power Electronics and Design
(ISLPED), pp. 1–6, 2019.
[40] A. Branco, L. Mottola, M. H. Alizai, and J. H. Siddiqui, “Intermittent asynchronous peripheral operations,” in Proceedings of the 17th Conference on Embedded Networked
Sensor Systems, SenSys ’19, (New York, NY, USA), p. 55–67, Association for Computing Machinery, 2019.
[41] A. Rahmati, M. Salajegheh, D. Holcomb, J. Sorber, W. P. Burleson, and K. Fu,
“TARDIS: Time and remanence decay in SRAM to implement secure protocols on
embedded devices without clocks,” in 21st USENIX Security Symposium (USENIX Security 12), (Bellevue, WA), pp. 221–236, USENIX Association, Aug. 2012.
[42] J. Hester, N. Tobias, A. Rahmati, L. Sitanayah, D. Holcomb, K. Fu, W. P. Burleson,
and J. Sorber, “Persistent clocks for batteryless sensing devices,” ACM Trans. Embed.
Comput. Syst., vol. 15, aug 2016.
[43] K. Ma, Y. Zheng, S. Li, K. Swaminathan, X. Li, Y. Liu, J. Sampson, Y. Xie, and
V. Narayanan, “Architecture exploration for ambient energy harvesting nonvolatile processors,” in 2015 IEEE 21st International Symposium on High Performance Computer
Architecture (HPCA), pp. 526–537, 2015.
[44] M. T. Islam and S. Nirjon, “Soundsemantics: Exploiting semantic knowledge in text
for embedded acoustic event classification,” in Proceedings of the 18th International
Conference on Information Processing in Sensor Networks, IPSN ’19, (New York, NY,
USA), p. 217–228, Association for Computing Machinery, 2019.
[45] Y. LECUN, “The mnist database of handwritten digits,”
http://yann.lecun.com/exdb/mnist/.
[46] P. Warden, “Speech commands: A dataset for limited-vocabulary speech recognition,”
2018.
[47] D. Arthur and S. Vassilvitskii, “k-means++: The advantages of careful seeding,” Technical Report 2006-13, Stanford InfoLab, June 2006.
[48] P. Black, “Dictionary of algorithms and data structures,” 1998-10-01 1998.
[49] B. George, “A study of the effect of random projection and other dimensionality reduction techniques on different classification methods,” Baselius Researcher, p. 201769,
2017.