臺灣博碩士論文加值系統

隨著科技不斷的進步,為了讓使用者生活更加便利,市面上各類智慧型裝置也 相繼出現,隨著IoT相關的產品興起,IoT這個領域也面臨到一個重大的問題。大 部分的產品之間都不具有操控互通性,因此使用者如果想要組合這些裝置產生更 複雜更強大的功能相當不容易。現在大部分的智慧型裝置都會有相對應專屬的手 機應用程式能去存取、控制裝置。因此我們想以逆向工程的角度切入,試著解決 這個問題。運用逆向工程的技術將手機應用程式中存取裝置相關的重要通訊資訊 取出,再將之轉成可以控制裝置的程式並提供統一個API來操控,如此便能達到 操控互通性。由於目前大部分的手機應用程式都有經過混淆來保護,所以想要取 出重要程式片段還先處理這個問題才行,因此本研究主要是在探討如何將混淆過 的Android應用程式反混淆。

With advences in technology, in order to make users’ life more convenient, many kinds of smart devices appear in the market. With IoT devices increasing, the domain face a major problem. Most of the products are not interoperable. Therefore, it is hard for users to compose their devices to more complex and powerful services. Nowadays, most of the smart devices have their own mobile applications which can access and control the devices. Thus, we try to achieve interoperability through reverse engineering. The idea is to use reverse engineering technology to extract the important communication information between mobile application and device; then generate control code which can control the device and provide a common interface to manipulate device. Because most of the mobile applications are obfuscated for protection, we want to extract the important program segment need to deal with this problem first. Hence, this thesis discusses about how we deobfuscate obfuscated Android applications.

誌謝 i
摘要 ii
Abstracts iii
List of Figures vii
List of Tables viii
Chapter 1 Introduction 1
Chapter 2 Related Work 5
2.1 DeGuard.................................. 5
2.2 Soot.................................... 6
2.3 Nice2Predict................................ 7
Chapter 3 Deobfuscation System 8
3.1 AndroidAPKSource........................... 8
3.2 LearningPhase .............................. 9
3.3 EvaluationPhase ............................. 9
3.4 PredictionPhase ............................. 9
Chapter 4 Dependency Grahp 11
4.1 ProgramElement............................. 11
4.2 ProgrammingRelation .......................... 12
4.3 Feature................................... 13 4.4 ConstraintandMethodGrouping .................... 14
4.5 FieldGrouping .............................. 14
4.6 RelationImprovement .......................... 14
Chapter 5 Programming Relation 15
5.1 Encapulation ............................... 15
5.1.1 Package .............................. 16
5.1.2 Class................................ 16 5.1.3 Modifier.............................. 17
5.1.4 Field................................ 18
5.1.5 MethodOperation ........................ 19
5.1.6 DataType............................. 23
5.2 Abstraction ................................ 23
5.2.1 Abstraction ............................ 24
5.2.2 Override.............................. 25
5.2.3 Overload.............................. 26
5.3 Delegation................................. 27
5.3.1 AccessedClass .......................... 28
5.3.2 AccessAttribute ......................... 29 5.3.3 Creation.............................. 30
5.3.4 InvokeMethod .......................... 30
5.3.5 InvokeConcreteMethod..................... 31
5.3.6 InvokeAbstractMethod ..................... 32
5.3.7 Delegation............................. 33
5.3.8 Intent ............................... 34
Chapter 6 Experimental Evaluation 36
6.1 ExperimentalEvaluation......................... 36
Chapter 7 Conclusion 38
Bibliography 39

[1]  Apache velocity engine. http://velocity.apache.org/.
[2]  Deguard. http://apk-deguard.com/.
[3]  F-droid. https://f-droid.org/.
[4]  Nice2predict. https://github.com/eth-srl/Nice2Predict.
[5]  B. Bichsel, V. Raychev, P. Tsankov, and M. Vechev. Statistical deobfuscation of android applications. In Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security, pages 343–355. ACM, 2016.
[6]  A. Einarsson and J. D. Nielsen. A survivor’s guide to java program analysis with soot.
[7]  J. Lafferty, A. McCallum, F. Pereira, et al. Conditional random fields: Probabilistic models for segmenting and labeling sequence data. 2001.
[8]  R. Vall ́ee-Rai, P. Co, E. Gagnon, L. Hendren, P. Lam, and V. Sundaresan. Soot - a java bytecode optimization framework. In Proceedings of the 1999 Conference of the Centre for Advanced Studies on Collaborative Research, CASCON ’99, pages 13–. IBM Press, 1999.