本文提出一個基於指數哥倫布編碼的輕量化且無損的資料壓縮法，適用於壓縮所收集到的感測器資料。由於之前的資料壓縮技術大部分都不是為了小型的嵌入式平台設計，其所需的運算資源巨大，故難以在資源十分有限的環境下執行。然而利用一種改良版的指數哥倫布編碼，可以在小型的嵌入式平台上，仍達到良好的壓縮效果。我們利用了這種改良版的指數哥倫布編碼，並增加了一個微型快取表的設計，以壓縮數種不同的感測器資料集。實驗結果顯示，我們所提出的方法在壓縮變化幅度不大的資料集的表現可以有些許的改善，且在那些變化較為劇烈的資料集，我們的小型快取表設計可以顯著的改善壓縮比。此外，我們的壓縮演算法仍然保持非常輕量化的特性，故非常適合使用在小型的嵌入式平台上。

This thesis describes a lightweight lossless compression algorithm based on exponential Golomb code. Such an algorithm is needed for compressing sensor data on sensor nodes, which are often too resource-constrained to run many common data compression algorithms. To address this problem, we propose an algorithm based on a variation of exponential Golomb code augmented with a tiny cache cache design to achieve good compression ratios. Experimental results show that our proposed algorithm is able to achieve competitive, often better compression ratios over different kinds of realistic sensor data with only modest computational resource demands.

Contents i
Acknowledgments vi
1 Introduction 1
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Thesis Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 Related Work 3
2.1 Lossless Compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.1 S-LZW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.2 LEC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Lossy Compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.1 K-RLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.2 LTC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3 Background Theory 6
3.1 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.1.1 Integers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.1.2 Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.1.3 Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.1.4 Strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2 Information Content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.3 Entropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.4 Source Coding Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.5 Golomb Code and Exponential Golomb Code . . . . . . . . . . . . . . . . . . . . . 9
3.5.1 Unary Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.5.2 Truncated Binary Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.5.3 Golomb Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.5.4 Exponential Golomb Code . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4 Technical Approach 13
4.1 Predictor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.2 Entropy Encoder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.3 Small Cache Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.3.1 Type One: Cache Table with a Hit/Miss Bit . . . . . . . . . . . . . . . . . . 16
4.3.2 Type Two: Cache Table without a Hit/Miss Bit . . . . . . . . . . . . . . . . 17
4.3.3 Type Three: Cache Table Combine with Truncated Binary Code . . . . . . . 18
4.4 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5 Implementations 24
5.1 Experimental Platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.1.1 EcoMini . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.1.2 Asthmagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.1.3 PC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.2 Firmware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
6 Evaluation 27
6.1 Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6.2 Experimental Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
6.2.1 Compression ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.2.2 Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.2.3 Energy and Power Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.3 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.3.1 Compression Ratio and Efficiency . . . . . . . . . . . . . . . . . . . . . . . 30
6.3.2 Space Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
6.3.3 Latency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6.3.4 Power Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6.4 Robustness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
6.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
7 Conclusions and Future Work 38
7.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
7.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Bibliography 40

[1] F. Marcelloni and M. Vecchio, “An efficient lossless compression algorithm for tiny nodes of monitoring wireless sensor networks,” The Computer Journal, vol. 52, no. 8, pp. 969–987, 2009.
[2] C. M. Sadler and M. Martonosi, “Data compression algorithms for energy-constrained devices in delay tolerant networks,” in Proceedings of the 4th international conference on Embedded networked sensor systems, pp. 265–278, ACM, 2006.
[3] M. Burrows and D. J. Wheeler, “A block-sorting lossless data compression algorithm,” 1994.
[4] E. P. Capo-Chichi, H. Guyennet, and J.-M. Friedt, “K-RLE: a new data compression algorithm for wireless sensor network,” in 2009 Third International Conference on Sensor Technologies and Applications, pp. 502–507, IEEE, 2009.
[5] T. Schoellhammer, B. Greenstein, E. Osterweil, M. Wimbrow, and D. Estrin, “Lightweight temporal compression of microclimate datasets,” 2004.
[6] G. Hansel, D. Perrin, and I. Simon, “Compression and entropy,” in Annual Symposium on Theoretical Aspects of Computer Science, pp. 513–528, Springer, 1992.
[7] C. Ji, L.-P. Chang, L. Shi, C. Gao, C. Wu, Y. Wang, and C. J. Xue, “Lightweight Data Compression for Mobile Flash Storage,” ACM Trans. Embedded Comput. Syst., vol. 16, pp. 183:1–183:18, 2017.
[8] D. J. MacKay and D. J. Mac Kay, Information theory, inference and learning algorithms. Cambridge university press, 2003.
[9] Embedded Platform Lab (EPL) at National Tsing Hua University (NTHU) in Taiwan, “Eco Mini.” http://epl.tw/ecomini/, 2014.
[10] “EFR32BG12 Blue Gecko Bluetooth ™ Low Energy SoC Family Data Sheet.” https://www.silabs.com/documents/login/data-sheets/efr32bg12-datasheet.pdf.
[11] “SensorScope deployments homepage.” https://lcav.epfl.ch/research/research-archives/research-archives-communications_and_sensor_networks_archive-html/sensorscope-en/page-145180-en-html/.
[12] G. B. Moody and R. G. Mark, “The impact of the MIT-BIH arrhythmia database,” IEEE Engineering in Medicine and Biology Magazine, vol. 20, no. 3, pp. 45–50, 2001.
[13] “CSS10 Chinese: Single Speaker Speech Dataset.”https://www.kaggle.com/bryanpark/chinese-single-speaker-speech-dataset.
[14] B. Koc, D. Sarkar, H. Kocak, and Z. Arnavut, “A study of power consumption on MSP432 family of microcontrollers for lossless data compression,” in 2015 12th International Conference on High-capacity Optical Networks and Enabling/Emerging Technologies (HONET), pp. 1–5,IEEE, 2015.