網路時脈偏移測量類似於網路校時，旨在利用每個封包的發送時間與接收時間所計算出來的偏移量來估計發送端與接收端的時鐘在時間行進速度上的差異。然而，現有的測量方法普遍只能對連續性分布的離群值進行測量。本篇論文則利用熵的特性，提出了一個更全面性的測量方法。對於偏移量隨著時間推移而呈現線性分布的有效測量數據，本方法皆能提供更穩定且更快速的測量。

在本研究中，將受測資料依據是否存在分段而分為兩類。在不存在分段的受測資料的測量結果中，本研究的方法與既有方法比較毫不遜色，測量變動範圍皆在1ppm以下。當存在分段時，既有方法的變動範圍有可能高達數十ppm以上，而本研究的方法的變動範圍仍然維持在1ppm以下。此外，本研究方法的時間複雜度為O(N)，對5000筆封包的測量時間僅需1.594秒。相較於其他方法，明顯地本研究的方法更有效率。

The process of clock skew measurement via network is to estimate the ticking rate difference between the sender and the receiver of packets by their offsets, or the receiving time subtracting by the sending time. Current approaches assume that the offsets form a straight-line-like cluster in the scatter diagram. This research, however, address skew measurement on offsets of multi-segment distributions. In this paper, an entropy-based method is developed to estimate the clock skews efficiently. As long as there exists a linear relationship between the offsets and the receiver’s time, the proposed method is able to provide stable and accurate estimation to both single-cluster and multi-segment cases.

In this study, offset-sets are divided into two types according to the existence of segmentation. For single-cluster case, the experimental results shows that the fluctuation range of estimated skew by the proposed method is bounded in 1 ppm, which is as good as the existing methods. When there exists breaks in an offset-set, however, the fluctuation range of skew may rise to dozens of ppm in other method, but skews derived by the proposed method still remain to have less than 1ppm fluctuation range. The time complexity of the proposed method is of only O(n), and the execution time for 5,000 packets is 1.594 seconds, which is obviously shorter than approaches like Hough transform-based method.

參考文獻
[1] Tadayoshi Kohno, Andre Broido, and Kc Claffy, “Remote Physical Device Fingerprinting,” IEEE Transactions on Dependable and Secure Computing, Volume 2, Number 2, Pages 93-108, 2005.
[2] Steven J. Murdoch, “Hot or not: Revealing Hidden Services by Their Clock Skew,” in Proceedings of the 13th ACM Conference on Computer and Communications Security, Pages 27-36, 2006.
[3] Sebastian Zander and Steven J. Murdoch, “An Improved Clock-skew Measurement Technique for Revealing Hidden Services,” in Proceedings of the 17th Conference on Security Symposium, Pages 211-225, 2008.
[4] Ding-Jie Huang, Wei-Chung Teng, Chih-Yuan Wang, Hsuan-Yu Huang, and Joseph M. Hellerstein, “Clock Skew Based Node Identification in Wireless Sensor Networks,” IEEE Global Communications Conference (GLOBECOM 2008), Pages 1-5, 2008.
[5] Suman Jana and Sneha Kumar Kasera, “On Fast and Accurate Detection of Unauthorized Wireless Access Points Using Clock Skews,”IEEE Transactions on Mobile Computing, Volume 9, Issue 3, Pages 104-115, 2008.
[6] Chrisil Arackaparambil, Sergey Bratus, Anna Shubina, and David Kotz “On the Reliability of Wireless Fingerprinting Using Clock Skews,” in Proceedings of the third ACM Conference on Wireless Network Security (WiSec 2010), Pages 169-174, 2010
[7] Ding-Jie Huang, Kai-Ting Yang, Chien-Chun Ni, Wei-Chung Teng, Tien-Ruey Hsiang, and Yuh-Jye Lee, “Clock Skew Based Client Device Identification in Cloud Environments,”the 26th IEEE International Conference on Advanced Information Networking and Applications , Pages 526-533, 2012.
[8] 鄭理介，《基於時脈偏移的可攜式裝置識別技術》，國立臺灣科技大學碩士論文，2013。
[9] Ding-Jie Huang and Wei-Chung Teng, “A Defense Against Clock Skew Replication Attacks In Wireless Sensor Networks,”Journal of Network and Computer Applications, Volume 39, Pages 26-37, 2014.
[10] Marius Cristea and Bogdan Groza, “Fingerprinting Smartphones Remotely via ICMP Timestamps,” IEEE Communications Letters, Volume 17, Issue 6,Pages 1081-1083, 2013.
[11] Swati Sharma, Alefiya Hussain, and Huzur Saran, “Experience with Heterogenous Clock-skew Based Device Fingerprinting,” in Proceedings of the 2012 Workshop on Learning from Authoritative Security Experiment Results (LASER 2012), Pages 9-18, 2012.
[12] Xiaowei Mei, Donggang Liu, Kun Sun, and Dingbang Xu, “On Feasibility of Fingerprinting Wireless Sensor Nodes Using Physical Properties,”IEEE 27th International Symposium on Parallel & Distributed Processing (IPDPS 2013), Pages 1112-1121, 2013.
[13] Md. Borhan Uddin and Claude Castelluccia, “Toward Clock Skew Based Wireless Sensor Node Services,”The 5th Annual ICST Wireless Internet Conference (WICON), Pages 1-9, 2010.
[14] Makoto Aoki, Eiji Oki, and Roberto Rojas-Cessa, “Measurement Scheme for One-Way Delay Variation with Detection and Removal of Clock Skew,” ETRI Journal, Volume 32, Number 6, Pages 854-862, 2010.
[15] Sue B. Moon, Paul Skell, and Don Towsley, “Estimation and Removal of Clock Skew from Network Delay Measurements,” in Proceedings of Eighteenth Annual Joint Conference of the IEEE Computer and Communications Societies (IEEE INFOCOM '99), Volume 1, Pages 227-234, 1999.
[16] Komang Oka Saputra, Wei-Chung Teng and Tsung-Han Chen, “Hough Transform Based Clock Skew Measurement Over Network,”IEEE Transactions on Instrumentation and Measurement, Volume 64, Issue 12, Pages 3209-3216, 2015.
[17] 楊祐齊，《一個改良的基於霍氏轉換時脈偏移測量方法》，國立臺灣科技大學碩士論文，2015。
[18] Paul V. C. Hough, “Method and Means for Recognizing Complex Patterns,”U.S. Patent 3,069,654, 1962.
[19] Richard O. Duda and Peter E. Hart, “Use of the Hough Transformation to Detect Lines and Curves in Pictures,”Communications of the ACM, Volume 15, Number. 1, Pages 11-15, 1972.
[20] Claude Elwood Shannon, “A Mathematical Theory of Communication,” Mobile Computing and Communications Review, Volume 5, Issue 1, Pages 3-55, 2001.
[21] Vern Paxson, “On Calibrating Measurements of Packet Transit Times,”ACM SIGMETRICS Joint International Conference on Measurement and Modeling of Computer Systems (SIGMETRICS 1998/PERFORMANCE 1998), Volume 26, Number 1, Pages 11-21, 1998.
[22] 周品言，《藍芽通訊傳輸延遲之研究》，國立臺灣科技大學碩士論文，2014。