設計穿戴式醫療裝置來監控與記錄資料，將會面臨缺乏有效動態功耗管理的挑戰。本論文提出一種功耗最佳化排程設計，運用於新型微控制器（microcontroller unit, MCU）、使用鏈結式直接記憶體存取（Linked DMA, LDMA）與即時作業系統（real-time operating system, RTOS）、整合多模態應用需求的實作方式，在不犧牲性能的前提下，得以最小化系統的運作功耗以及CPU所需執行時間。LDMA與傳統DMA不同之處，在於能夠獨立自主處理一連串的輸入、輸出操作，得以節省主處理器的執行指令數量，藉此獲得更多進入節能模式的機會。實驗數據顯示，相較於傳統DMA的設計方式，使用LDMA可以讓主處理器的使用率減少36%，進而節省4%的整體能源消耗。

A challenge with designing wearable medical devices is that they need to be monitoring and logging data and cannot be easily power managed by dynamic power management (DPM) techniques. This thesis proposes a power-mode scheduling method on the microcontroller unit (MCU) for multimodality sensing while taking advantage of the novel “linked direct-memory access” (LDMA) on modern MCUs and Micrium OS to minimize power consumption and execution time by the processor core while maintaining the required performance. Unlike traditional DMA, LDMA enables the controller to autonomously handle a sequence of transactions that would require processor instructions to control in traditional DMA. Experimental results show a reduction of 36% CPU utilization, which results in 4% energy savings in system-level compared to baseline DMA.

Contents i

Acknowledgments vi

1 Introduction 1
1.1 Motivating Application: CPR Data Logging . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Bluebox System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Thesis Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2 Background and Related Work 4
2.1 Background on I/O for Embedded System . . . . . . . . . . . . . . . . . . . . . . . 4
2.1.1 Program I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1.2 Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1.3 Direct Memory Access (DMA) . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1.4 Linked DMA (LDMA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2 Low-Power Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.1 Power Management on MCUs . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.2 Power Management for Peripherals . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Body Sensor Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 Real-Time Operation System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3 Illustrative Example 11
3.1 Application: Medical Data Logging . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.1.1 Background: Code Blue . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.1.2 Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.2 System Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.2.1 MCU Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.2.2 Bluebox Peripheral Components . . . . . . . . . . . . . . . . . . . . . . . . 15
3.3 Application Workload . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.4 Data Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.4.1 ADS1298 Data Buffer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.4.2 ICM20948 Data Buffer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.4.3 Page Arrangement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

4 Linked DMA 19
4.1 Background of Linked-DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.2 Peripheral Reflex System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.3 Iterative routine processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.4 Comparison between Program I/O, DMA, LDMA . . . . . . . . . . . . . . . . . . 24
4.5 Hardware Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.5.1 Peripheral Reuse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.5.2 Protocol Pin Reuse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.6 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

5 Real-Time Operating System 28
5.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.2 Bare-Metal vs. RTOS base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.3 Impact of RTOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.4 Remaining Time Improving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.5 Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.5.1 Real-Time Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.5.2 Priority Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.5.3 Schedulability Proving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.6 Scheduling Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

6 Evaluation 34
6.1 Power Measuring Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
6.2 Hardware Limitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
6.2.1 Sampling Mechanism of ADS1298 and ICM20948 . . . . . . . . . . . . . . 34
6.2.2 ICM20948 Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
6.3 Multi-Modality Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.3.1 Cooperation between Sensors . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.3.2 Duty Cycle Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.3.3 Load Balancing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.4 Peripheral to Peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
6.4.1 SPI to SPI Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
6.5 Power Mode Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
6.5.1 Switch Mode Considered . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
6.5.2 Test Current Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
6.6 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
6.6.1 Overhead proportion: CPU Time . . . . . . . . . . . . . . . . . . . . . . . . 42
6.6.2 Overhead proportion: Current . . . . . . . . . . . . . . . . . . . . . . . . . 42
6.6.3 Comparison of Related Works . . . . . . . . . . . . . . . . . . . . . . . . . 43

7 Conclusions and Future Work 45
7.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
7.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Appendix 50
Photoplethysmography Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Pulse Width Modulation vs. Interrupt Request . . . . . . . . . . . . . . . . . . . . . 50
Light Intensity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
SPI Serial Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

[1] C. Kac, C. Lee, Y. Yeh, J. Sung, and P. H. Chou, “Ecosim: A smartphone-based sensor-node emulator with native sensors and protocol stack,” in 2019 International Symposium on VLSI
Design, Automation and Test (VLSI-DAT), pp. 1–4, 2019.

[2] T. Enami, K. Kawakami, and H. Yamazaki, “Dma-driven control method for low power sensor node,” in 2015 IEEE Topical Conference on Wireless Sensors and Sensor Networks (WiSNet), pp. 53–55, 2015.

[3] F. Reverter and M. Gasulla, “Experimental characterization of the energy consumption of adc embedded into microcontrollers operating in low power,” in 2019 IEEE International Instrumentation and Measurement Technology Conference (I2MTC), pp. 1–5, 2019.

[4] Chou, Ortega, and Borriello, “Synthesis of the hardware/software interface in microcontrollerbased systems,” in 1992 IEEE/ACM International Conference on Computer-Aided Design,
pp. 488–495, 1992.

[5] Y. Lin, P. Hsiu, and T. 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.

[6] A. Pullini, D. Rossi, G. Haugou, and L. Benini, “udma: An autonomous i/o subsystem for it end-nodes,” in 2017 27th International Symposium on Power and Timing Modeling, Optimization and Simulation (PATMOS), pp. 1–8, 2017.

[7] A. Heydarigorji, S. M. Safavi, C. T. Lee, and P. H. Chou, “Head-mouse: A simple cursor controller based on optical measurement of head tilt,” in 2017 IEEE Signal Processing in Medicine and Biology Symposium (SPMB), Dec 2017.

[8] F. Reverter, “Toward non-cpu activity in low-power mcu-based measurement systems,” IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 1, pp. 15–17, 2020.

[9] X. Zhang, H. Jiang, B. Zhu, X. Chen, C. Zhang, and Z. Wang, “A low-power remotelyprogrammable mcu for implantable medical devices,” in 2010 IEEE Asia Pacific Conference on Circuits and Systems, pp. 28–31, 2010.

[10] M. Hayashikoshi, H. Noda, H. Kawai, Y. Murai, S. Otani, K. Nii, Y. Matsuda, and H. Kondo, “Low-power multi-sensor system with power management and nonvolatile memory access control for iot applications,” IEEE Transactions on Multi-Scale Computing Systems, vol. 4, no. 4, pp. 784–792, 2018.

[11] Y. Lin, S. Lee, Y. Chen, C. Huang, and H. Chiueh, “Live demonstration: A wireless multichannel physiological signal acquisition system-on-chip for wearable devices,” in 2016 IEEE
Biomedical Circuits and Systems Conference (BioCAS), pp. 128–128, 2016.

[12] A. K. Vishwakarma, K. V. Suresh, and U. K. Singh, “Porting and systematic testing of an embedded rtos,” in International Conference on Computing and Communication Technologies, pp. 1–4,
2014.

[13] F. A. Mohideen, “Rtos for pic18 microcontrollers,” in 2010 5th International Conference on Industrial and Information Systems, pp. 275–280, 2010.

[14] “Nyquist–shannon sampling theorem.” https://en.wikipedia.org/wiki/Nyquist\T1\textendashShannon_sampling_theorem.

[15] C. Jácome and A. Marques, “Computerized respiratory sounds in patients with copd: a systematic review.”https://www.ncbi.nlm.nih.gov/pubmed/24914587.

[16] O. D. Lara and M. A. Labrador, “A survey on human activity recognition using wearable sensors,” IEEE Communications Surveys Tutorials, vol. 15, no. 3, pp. 1192–1209, 2013.

[17] AN0025: Peripheral Reflex System (PRS).

[18] A. Silberschatz, G. Gagne, and P. B. Galvin, Operating System Concepts. Wiley Publishing, 8th ed., 2011.

[19] C. Liu and J. Layland, “Scheduling algorithms for multiprogramming in a hard real-time environment,” J. ACM, vol. 20, pp. 46–61, 1973.

[20] EFR32BG12 Gecko Bluetooth ®Low Energy SoC Family Data Sheet.