Forensics Analysis of Privacy of Portable Web Browsers
Forensics Analysis of Privacy of Portable Web Browsers
Part of the Aviation Safety and Security Commons, Computer Law Commons, Defense and Security
Studies Commons, Forensic Science and Technology Commons, Information Security Commons,
National Security Law Commons, OS and Networks Commons, Other Computer Sciences Commons, and
the Social Control, Law, Crime, and Deviance Commons
ABSTRACT
Web browser vendors offer a portable web browser option which is considered as one of the
features that provides user privacy. Portable web browser is a browser that can be launched from
a USB flash drive without the need for its installation on the host machine. Most popular web
browsers have portable versions of their browsers as well. Portable web browsing poses a great
challenge to computer forensic investigators who try to reconstruct the past browsing history, in
case of any computer incidence. This research examines various sources in the host machine such
as physical memory, temporary, recent, event files, Windows Registry, and Cache.dll files for the
evidential information regarding portable browsing session. The portable browsers under this study
include Firefox, Chrome, Safari, and Opera. Results of this experiment show that portable web
browsers do not provide user-privacy as they are expected to do.
Keywords: computer forensics tools, RAM forensics, volatile memory, forensics artifacts,
Registry
between the portable web browsers with which creating these tools can be found in Redline
we experimented. user manual.
Table 3 The experiment consisted of two parts. In
SQLite report of portable browsing session the first part, after a portable browsing
Portable Suspect machine
session, we left the portable browser flash drive
Browser Activity
attached to the suspect machine, captured, and
Google Chrome cookies.sqlite-wal, places.sqlite-shm,
analyzed the RAM. In the second part, we
and webappsstore.sqlite-shm were
deleted
removed the portable browser flash drive,
profile/*.db were modified captured the RAM, and analyzed it. RAM
Firefox cookies.sqlite-wal, places.sqlite-shm, capture in the latter part is very time sensitive
and webappsstore.sqlite-shm were and it depends on the time gap between
deleted removal of portable browser flash drive and
profile/*.db were modified RAM capture. Since Redline Collector cannot
Safari cookies.sqlite-wal, places.sqlite-shm, collect information about terminated processes
and webappsstore.sqlite-shm were and closed files, we also used WinHex (2015)
deleted Hexadecimal editor.
profile/*.db were modified
Opera cookies.sqlite-wal, places.sqlite-shm, RAM Forensics Process
and webappsstore.sqlite-shm were
To make data extracting less cumbersome, we
deleted
profile/*.db were modified cleared all cookies, cache, history, bookmarks,
etc that may have been left on the suspect
For RAM forensics we followed the machines from our earlier experiment. We
framework suggested by Ghafarian (2015). We installed Memoryze software on the forensics
chose an open source memory forensics tool workstation. To simplify analysis, we disabled
called Redline (2015) for the following reasons: physical address extension mode on Redline.
We ran Redline, created the RAM capture
Graphical User Interface software Collector, and saved it on a wiped
Selection option which allows user USB external drive. Then we followed the
to choose only browsing related below steps:
processes and disabling all the other 1. Attached the portable browser flash
processes and files drive to the suspect machine and
Allow to import memory analysis configured the browsers as the default
results to a file such as MS Word browser with extensions and plug-ins
for offline processing disabled. Then we performed a
browsing session, attached the Collector
Easy to user and having a external drive to the suspect machine,
comprehensive user manual. captured RAM, saved the file onto the
In Redline, the RAM capture tool is called external drive and removed external
‘Collector’ and the RAM analysis is called drive for RAM analysis.
‘Memoryze.’ We created the Collector software 2. Step 1 was repeated for all the other
and saved it on a USB external drive. We also suspect machines with different
created the Memoryze and saved it on the portable browser.
Forensics workstation machine. The details of 3. We repeated steps 1 and 2 above, but
this time we removed the portable
browser flash drive and immediately suspect machines and immediately RAMs were
captured the RAM and saved it to the captured.
external drive.
The blue entries in Table 4 show that with
4. Configured Memoryze to retrieve only
the exception of email password everything else
browsing-related information and
was retrievable. That means that if the
processes. This action reduced the
portable USB flash drive is attached to the
amount and time of data analysis. We
machine during RAM capture, portable
imported the memory parsed data to a
browser provides no privacy at all. In this case,
MS Word file for offline analysis. We
the information that was retrieved from
should note that Redline only provide
memory is enough to conclude browsing
information about running processes
activities and establishing link between the
and programs that were running before
web browsing activities and the suspect. For
memory was captured. We also used
example, browsing history, search history, and
WinHex to retrieve residual data on
file downloads were retrieved from memory for
these processes and files.
all of the portable browsers we studied. These
5. Step 4 was repeated for the other three
are important evidential information for
captured RAM files.
computer-forensics investigators.
Over all, we had four RAM captured files
However, when we removed the portable
for the cases when portable browsers flash
browser from the PC and then captured RAM,
drives were still attached to the suspect
the forensics artifacts retrieved from main
machine during the RAM capture process. And
memory slightly varies among various
four captured RAM files for the case when
browsers. This variation is discussed below.
portable storage flash drives were removed
after each browsing session. The total memory For Mozilla Firefox, analysis of the
files that we captured were eight. Considering memory dumped file showed considerable
each RAM capture took on average one hour, browser-related entries in memory indicating
we spent eight hours to capture the memory of web browser activity. We were able to detect
the suspect machines. The process of memory email communication details, browsing and
capture and analysis were performed according URL history, search history, and downloaded
to the forensics investigations rules and files (documents, images, and videos). On one
regulations. The results are discussed in the hand, some information such as cookies, email
next section. password, timelines, and process ID could not
be retrieved. We also used Winhex to analyze
RESULTS the captured RAM, but did not find any of
aforementioned data. This indicates that when
Retrievable computer forensics artifacts after a the portable browser flash drives were
portable browsing session via memory forensics removed, some of the browsing information
for all four browsers are summarized in Table was removed from the memory.
4. The blue columns represent artifacts Similar results were observed for Opera.
retrieved when the portable browsers flash Analysis of the RAM showed that portable
drives were still attached to the machines browser flash drive removal had an effect on
when RAMs were captured. The pink columns the amount of data retrievable from memory.
represent retrieved data when the portable Similar to Firefox cookies, timelines, email
browser flash drives were removed from the passwords, and process ID were deleted before
we captured RAM. This is because once the Chrome portable left the most residual
portable browser flash drive is removed, the artifacts on the host machine.
data structure tree that handles cookies, for
For Safari, the amount of retrieved data
example, are not accessible. On the other
from portable browsing session is identical to
hand, we were able to identify HTML data
Firefox and Opera. That means cookies,
containing various types of information
timeline, and email password were not
including the SSL certificate for accessing a
retrievable from main memory. However, like
secure website, URL, file downloaded and
Firefox, we were able to see forensically-
more.
valuable information such as history, file
Analysis of physical memory for portable downloads, SSL certificates, etc.
Google Chrome revealed forensically valuable
In an attempt to validate the retrieved
artifacts such as Certificate, HTML text file,
data from main memory, we used another open
URL history, and files downloaded. We should
source software tool from Microsoft called
note that only Google Chrome saved process
DumpIt to capture the physical memory after
ID in memory. Nevertheless, similar to Firefox,
a browsing session. Then we used WinHex to
we were not able to see cookies, email
analyze the captured images. The results for
passwords, and timeline. Analysis of the
both Redline and DumpIt were identical.
retrieved artifacts indicates that Google
Table 4
Results of Installed Browser, and Portable Browsers
Data Item Firefox Opera Chrome Safari Firefox Opera Chrome Safari
Removed Removed Removed Removed Attached Attached Attached Attached
Browser - - √ - √ √ √ √
process
URL History √ √ √ √ √ √ √ √
Cookies − − − − √ √ √ √
File downloads √ √ √ √ √ √ √ √
Timelines − − √ − √ √ √ √
Browser √ √ √ √ √ √ √ √
history
Email − − − − − − − −
password
Email ID √ √ √ √ √ √ √ √
SSL Certificate √ √ √ √ √ √ √ √
Search history √ √ √ √ √ √ √ √
Figure 3. Memory analysis reveals the date and time when the user access the site
FUTURE WORK
This research can be extended in several ways;
first, determine better tools and methodologies
REFERENCES
Adautin, E.D. and Meeran, N. (2015). Forensic through Memory Forensics. International
Reconstruction and Analysis of Residual Journal of Computer Applications (0975 –
Artifacts from Portable Web Browse. 8887), 132(16), 27-34
International Journal of Computer
Hejazi, S.M., Talhi, C. & Debbabi, M. (2009).
Applications (0975 – 8887), 128 (18), 19-24.
Extraction of Forensically Sensitive
Aggarwal, G., Bursztien, E., Jackson C., & Information from Windows Physical
Boneh, D. (2010). An analysis of private Memory. Digital Investigation, 6, 121-131.
browsing modes in modern browsers. Elsevier publishing Co.
Proceedings of the 19th Usenix Security
Koepi, D. (2010). Firefox Forensics. Retrieved
Symposium.
November 2014 from
Amari, K., (2009). Techniques and Tools for http://davidkoepi.wordpress.com/2010/11/
Recovering and Analyzing Data from 27/firefoxforensics
Volatile Memory. SANS Institute InfoSec
Mahendrakar, A., Irving, J., and Patel, S.,
Reading Room.
(2010). Forensic Analysis of Private
Choi, J. H., K.G. Lee, J. Park, C. Lee, and S. Browsing Mode in Popular Browsers.
Lee (2012). Analysis framework to detect Retrieved August 2014 from
artifacts of portable web browser. Center http://mocktest.net/paper.pdf
for Information Security Technologies
Mandiant Redline User Manual (2014).
DaemonFS, Retrieved on May 27 from Retrieved February 2015
http://sourceforge.net/projects/daemonfs/ fromhttps://dl.mandiant.com/EE/library/
Redline1.7_UserGuide.pdf
Davis, N. (2009). Live memory forensics for
Windows Operating Systems. Eastern Marringhton, A., I. Baggili, T. AI Ismail, A.
Michigan University, IA 328. Retrieved, AI Kaf (2013). Portable Web Browser
January 2015 from Forensics: A forensic examination of the
https://www.emich.edu/ia/pdf/research/Li privacy benefits of portable web browsers.
ve%20Memory%20Acquisition%20for%20W IEEE Journal.
indows%20Operating%20Systems,%20Naja
NirSofer. (2013). NirSoft Freeware Utilities.
%20Davis.pdf
Retrieved February 2015 from
Dharan, D.G. and Meeran, N.A.R. (2014). http://nirsoft.net
Forensic Evidence Collection by
Oh, O., Lee, S., and Lee, S. (2011). Advanced
Reconstruction of Artifacts in Portable
evidence collection and analysis of web
Web Browser. International Journal of
browser activity. Journal of digital
Computer Applications (0975 – 8887),
investigation 8, 62-70
91(4), 32-35.
Ohana, D.J. and Shashidhar, N. (2013). Do
Ghafarian, A. and Hosseini, S.A. (2015).
private and portable web browsers leave
Analysis of Private Browsing Mode
incriminating Evidence?: a forensic analysis