Android anti-forensics through a local paradigm

By Alessandro Disfano, Gianluigi Me, Francesco Pace

11/08/2013 Fri.Daun Jeong

Introduction

Definition of Anti-forensics

The Android Operating System

Android Anti-forensics

Experiments

Conclusion

Anti-forensics techniques applied to mobiles devices

Test for effectiveness of such techniques vs. both the cursory examination of the device and some acquisition tools

Trend Uptick in the use of Anti-forensics Confined in the classic forensics

environment⇒ The instance of some common AF techniques to Android mobile devices

Any attempts to compromise the availability or usefulness of evidence in the forensic process. The availability of evidence can be

compromised by preventing its creation, hiding its existence and by manipulating the evidence.

The usefulness can be compromised by deleting the evidence or by tampering its integrity.

1. Destroying Evidence: Destruction of evidence in order to make it unusable.

2. Hiding Evidence: Decreasing the visibility of the evidence

⇒ Both 1 and 2 processes can make other evidence

3. Eliminating Evidence Sources: Preventing Evidence Creation

4. Counterfeiting Evidence: Creating a fake version of evidence.

Application & Sandboxes: Android binds any running application to

a secure Sandbox which cannot interfere with any other application.

User IDs and Permissions: Android manages each application as a

different Linux user. Includes <uses-permission> tags in

application’s Android manifest.xml.

1. Current Android Forensics Techniques & Tools

2. Instantiating Anti-forensics

3. The Evidence Export Process

4. The Evidence Import Process

5. The Evidence Destruction Process

1. Android Debug Bridge (ADB)A tool provided with Android SDK which allows

the interaction between the mobile device and a remote station.

2. Nandroid BackupNandroid is a set of tools supporting the

backup and restore capabilities for rooted Android devices.

Support the full NAND flash memory imaging which can be performed by a special boot mode.

3. Physical Imaging by ddThe dd tool allows the byte-level physical

imaging of Unix files and can be applied to regular files and to devices files as well because of the availability of a Unix-like command shell.

4. Commercial ToolsCommercial Tools: Parabon corporation, Micro

Systemation, Celle brite.Open Source Tools: Mobile Internal Acquisition

Tool.

5. Serial Commands over USBCapabilities to eavesdrop the data

conveyed over-the-wire.

6. Simulated SD cardTo use a modified update file in order to

avoid the destruction of internal memory data and to provide kernel-level tools to support the acquisition of data.

7. Softeware ApplicationApplications that are able to explore, read,

and mirror the contents stored by the File System even for the internal memory storage volume.

Exploiting android features Strong Linux process & User management

policies

A private folder : A directory that is inaccessible for any other applications Private folders in internal memory are hard to examine

because of isolation and physical imaging

Anti-forensics by a common application: Evidence Export/Import/Destruction Process

1) Android Destroying Evidence: Text messages, The browser bookmark, Call log⇒ Deletion of Related Database

2) Android Hiding Evidence: Multimedia files ⇒ Move them into internal storage (private folder)

3) Android Eliminating Evidence Sources: Multimedia Messages (MMSs)⇒ Modify identifiers to be invisible to end-user

4) Android Counterfeiting Evidence: Contact Information⇒ Modify flag & related number

Restore the previous state of the device.

The private storage of the evidence Organize the exported evidence using set of

common files in the private folder A XML-style file(export.xml) is responsible

for the storage of all evidence A number of files of various format are

imported by the removable memory card.

How to reconstruct the evidence?

Fully Automated Evidence Reconstruction: AFDroid1) Private folder inspection2) export.xml file processing▪ Related DB & table▪ The connection DB

3) Other file processing

Internal Memory & Data Recovery It is still incomplete to acquire the image of

internal memory. (JTAG)

Fully Automated Process⇒ Uninstall of AFDroid All the related data are logically deleted by

the FS. Can avoid human errors. Reduces time.

Objectives : To test the strength of the Evidence

Export/ Destruction process in relation to the tools that are currently able to acquire a snapshot of the internal memory of the target device.

Used devices : Samsung Galaxy i7500 device equipped

with the Android 1.5 S 아 . Used acquisition tools

: Paraben Device Seizure/Nandroid/MIAT

Experimental Workflows1) Evidence export process▪ First imaging with Nandroid tool▪ Execution of AFDroid▪ Acquisition with MIAT tool▪ Second imaging with Nandroid tool

2) Evidence destruction process▪ First imaging with Nandroid tool▪ Execution of AFDroid▪ Second imaging with Nandroid tool▪ Uninstall of the AFDroid▪ Acquisition with the MIAT tool▪ Third imaging with the Nandroid tool

Cursory examination of the SMS/MMS database before and after the EEP. The entire se of SMS/MMS message is emptied .

The Nandroid tool and MIAT tool can recover all the evidence that has been previously exported in the private folder

A large amount of the multimedia data can negatively affect the duration of the process.

It is realistic to suppose that just reduced amount of such data can be exported into the private folder because of the limited capacity of the current internal memory.

When the application is uninstalled and the EDP completed, private folder is removed including all the stored contents.

After that, neither the Nandroid nor the MIAT tools were able to recover the deleted data.

Current and Future Work 1. Improving the AFDroid application▪ To selectively choose the target evidence▪ The expansion of the kinds of target evidence

2. Expanding the compatibility to other operating system ▪ Windows Moble, Symbian.

Android Anti-Forensics Through a Local Paradigm. Alessandro Distefano, Gianluigi Me and Francesco Pace, Digital Investigation 7 (2010) s83-s94.