Client Forensics: An Assessment of Existing Research And Future Directions

Rose Shumba
International Journal of Computer Science and Security (IJCSS), Volume (12) : Issue (3) : 2018 45
Client Forensics: An Assessment of Existing Research And
Future Directions
Rose Shumba shumba@usna.edu
Cyber Science Department
United States Naval Academy
Annapolis, MD
Abstract
This paper provides an assessment of processes for identifying artifacts, left on client devices
after a cloud storage interaction. It focuses on those artifacts that may be used to prove usage of
a cloud service, as proposed by the current research. Besides providing the current state of
knowledge in client forensics, this paper (1) provides a summary of current research in the area
of client forensics, (2) presents similarities and differences among proposed processes and
identified artifacts, and (3) presents some possible future work. Investigators need to understand
how devices and cloud storage services interact, the types of evidential artifacts that are likely to
remain on the devices after cloud storage interaction, and how they may be used to prove usage.
Not knowing if a cloud service was accessed, or which cloud service or the location of digital
evidence can potentially impede an investigation.
Keywords: Cloud Forensics, Cloud Storage Services, Client Forensics.
1. INTRODUCTION
A cloud can be defined as a scalable, virtualized, distributed computing platform, whose shared
resources are accessed remotely by users through a network. There are three primary cloud
service models; Software as a Service (SaaS), Platform as a Service (PaaS), and Infrastructure
as a Service (IaaS). The fourth delivery service model, which has emerged because of the ever-
increasing cloud storage options, is Storage as a Service (StaaS). StaaS has been widely
adopted by governments, organizations, and individuals, with Gartner forecasting that a third of
user data will be stored in the cloud[1]. Each service model mentioned above is deployable as a
public, private, hybrid or community. A description of the service and deployment models is
beyond the scope of this paper.
As the use of cloud services continues to transform IT Services by providing benefits such as
increased flexibility, efficiency, and costs, the security of corporate data is becoming a concern.
Cyber-crime, such as theft of intellectual property, espionage, acceptable use policy violations,
and data breaches is on the rise. When such incidents occur, it may be necessary to conduct
investigations. The National Institute of Standards and Technology [2] and several researchers
identified over 65 challenges associated with cloud forensics.[3]; [4,5,6]; [8]; [9]; [10],
Depending on the deployment architecture and service model, it is possible that investigators
may not have access to the physical servers to conduct server analysis making it hard to
determine the legal framework to follow to obtain the evidence. It leaves the investigator with
three options:
1) To attempt and recover evidence from seized local devices known to have interacted with
the cloud,
2) To attempt to eavesdrop network traffic between local devices and the cloud network,
and

Rose Shumba
3) To request a court in a foreign jurisdiction to seize evidence directly from a cloud
server[3].
The last option brings additional legal challenges, such as the problem of identifying and
addressing issues of jurisdiction for legal access to data and the lack of adequate channels for
international communication and cooperation during cyber forensic investigations. Section 2 of
this paper provides a summary of the research work, including the experiments carried out by the
different researchers, similarities, and differences between processes used and identified
artifacts, section 3 presents further questions for future research.
2. LANDSCAPE OF EXISTING RESEARCH WORK
Unique to client forensics analysis, research has focused on identifying remnants that can be
used to prove usage. A cloud storage service is accessible through installed client software or a
browser. Researchers carried out some experiments that involved accessing a storage service
from one or multiple devices:
1) Google Drive and Dropbox from a Windows 7 PC and an iPhone 3G [4,5,6]
2) Amazon S3, Dropbox, Evernote, and Google docs from Windows XP/Vista/7, a Mac PC, and
an iPhone 3G [7].
3) Dropbox, Google Drive, and SkyDrive from a Windows 7PC and iPhone 3G [9].
4) Copy and ownCloud from Windows 8.1 PC[8].
5) Google Documents, Flickr, PicasaWeb, Dropbox from a Windows 7 PC [3]
6) 360 and Baidu from a Windows 7 PC[10]
Researchers used a combination of static and dynamic processes to identify the artifacts. The
static approach assumes that the investigator has a forensic image and can use forensic tools
and prior acquired knowledge and skills to locate remnants [8]. The dynamic method uses
software tools, such as Disk Pulse and RegShot to find artifacts[9], while the experiment activity
is underway and the PC being used to access the service is on.
Table 1 shows for each researcher, the cloud storage service, and the devices used to access
the service. Also, the PC platform, in order of popularity and associated browsers is presented.
Mobile access used an iPhone 3G with inbuilt Safari browser.
The experiments core activities involved installing, accessing cloud service through a web
browser and the client software, involved uploading/copying and deleting the user data files.
From Table 1, it is clear that most the Windows 7 using IE browser has been the commonly used
platform and the 3G iOS device. This shows the areas of future experiments.
2.1 Conclusions From The Experiments Carried Out – Current Research
A closer look at the processes and identified artifacts, from the experiments described in the
previous section, shows the following:
1) Accessing a cloud storage service through a Web browser or client software creates a
substantial amount of artifacts that can be used to prove usage of the service. Examples
of remnants include the cloud storage service used, installation location, installed version,
usernames, passwords, URLs of downloaded client software, prefetch files, link files, and
file references related to the account. These artifacts play a significant role in an
investigation as they may lead an investigator to the potential location of other remnants
promptly.
2) The identified residues are dependent on the browser used to access the storage. From
the experiments, we can see that Ephani experimented with Mozilla Firefox and Internet
Explorer[9], Chung used Internet Explorer[7], and Quick, and Choo used Mozilla Firefox,
Google Chrome, Safari and Internet Explorer with Dropbox and Google Drive[4,5,6].

Rose Shumba
3) Quick and Choo's found a Google Drive account username through browser analysis with
Mozilla and Google Chrome, but none with Apple Safari. For Dropbox, the account
username was identified from browser history with Mozilla Firefox and Google Chrome
browsers.
4) The principal sources of artifacts depend on the device used to access the storage.
Where a PC is used the three sources of remnants are the hard drive, the RAM and the
eavesdropped network traffic between the device and the cloud network. When an
iPhone 3G the specific locations for artifacts were database files, XML files, and plist
files, as identified from the logical extract.
5) The different experiments, using the same cloud service and device, identified similar
artifacts. Tables 2 and 3 compare identified remnants when Dropbox and Google Drive
are accessed from a PC as presented by Ephani[9] and Quick [4]). The experiment
activities involved downloading the client software for installation, performing various
operations including uploading and deleting data files, accessing storage service through
a browser and client software and uninstalling the service application. For example, if a
suspect installed the Google Drive client software, the investigator can look in the
Program Files folder for proof of installation.
6) With Google Drive and Dropbox, Choo [4,5] showed that additional artifacts to prove
usage are obtainable from the network traffic analysis. These include the IP address,
registered owners, and the digital certificates. When Internet Explorer was used to
access Google Drive, the username was observed in the network traffic. RAM analysis
after accessing drive through a browser also showed the username.
7) Except for Choo [4,5] and Long [10], the presented research did not follow a process.
Choo and Long [10] proposed methods for client forensics analysis on a PC client. Choo
[4,5] developed a standard analysis process for practitioners, examiners, and
researchers to follow for client forensics. The proposed method can be used by an
investigator to find out Google Drive data remnants when accessed from a local device.
The process is cyclic and consists of the following phases:
a. setting the scope of the investigation,
b. setting up the virtual machines, installing the tools and browsers,
c. identifying and collecting the virtual hard drive image, memory image, and
network capture file,
d. making forensic copies of each image,
e. using a range of forensic tools to analyze each image,
f. stopping the investigation, or further analysis, if required.
Long developed a process that involved analyzing the registry on the user device, collecting
evidence from the browser and client and then determining the possible event sequence by
examining the correlations and rules of user activities among different time, different targets, and
behavior intention. The process was tested with 360 cloud and Baidu cloud services.
2.2 Further Questions And Future Work
The main conclusion from the current research work is that some evidentiary artifacts that can be
used to prove usage are obtainable through an exhaustive analysis of the client devices, without
accessing the cloud server.
Recovering evidence from client devices known to have interacted with the cloud has the
advantages that the devices can easily be accessed, and the cost of the forensic analysis is
relatively low.
Depending on the deployment model in use, much of the evidence may be on the client. An
exhaustive analysis of the client devices may help us obtain details, such as account credentials
that can be used to access the server side. Location of evidence and having precise knowledge
of what information is on the device can expedite current casework.

Rose Shumba
There are further questions that need to be addressed as far as client forensic analysis research
is concerned.
1) Presented study limits the scope of an investigation to a single device. How do we merge
the artifacts/evidence from several devices belonging to a single user into a timeline of
events to gain an accurate depiction of the user activities?
2) How can the identified artifacts be used to speed up an investigation?
3) How can this work be extended to include more commonly used mobile device platforms;
Android, Windows and latest iPhone?
4) How can artifacts gather at the client side be used to access account data on the server
side?
5) What is the relationship between the remnants collected on the client side and those on
the server side?
6) Does the version of the client software matter as to the types of identified artifacts?
A fundamental observation is that software developers are continually providing additional
functionality, closing security holes; this affects how much remnants are collectible from the client
side. For example in 2012, Chung [7] research identified a database file (file cache.db) in
Dropbox, as being of high value to the investigator, since it contained historical synchronization
logs. In 2013, Choo [4] and Ephani[9] showed that the file was encrypted and a unique tool was
now needed to decrypt it. Also, several different storage services apps are now available, making
it a challenge to collect remnants for each. How can a more unified digital investigation process
be developed to cater for different client software? Is there a better way of investigating storage
services?
3. CONCLUSION
The growing popularity of cloud storage services means that this media will be used for
cybercrime, resulting in more investigation cases. One challenge is maintaining a chain of
custody in the cloud. There is a need or more research in this area.
Future work should include accessing popularly used cloud storage services from commonly
used mobile platforms; Android, Windows, and the latest iPhone. A series of experiments that
involve installing, accessing, uploading and downloading some documents; uninstall the client
software, and then using anti-forensics techniques (deletion, uninstalling and clearing the browser
history) have been performed for the popularly used cloud services, accessed from Android. For
each mobile device, a physical extract will be used. We have started work on rooting an Android
device, in order to collect the main system folders.
4. REFERENCES
[1] Cloud forensics definitions and critical criteria for cloud forensic capability: An overview of
survey results. Digital Investigation, 10(1), pp 34-43 [2016].
[2] Stamford. (n.d.). Gartner Says That Consumers Will Store More Than a Third of Their Digital
Content in the Cloud by 2016. Available: http://www.gartner.com/newsroom/id/2060215
[3] NIST. (2014). NIST Cloud Computing Forensic Science Challenges (NISTIR 8006).
[4] Marturana, F., Me, G., & Tacconi, S.. A Case Study on Digital Forensics in the Cloud. In
Proceedings of the 2012 International Conference on Cyber-Enabled Distributed Computing
and Knowledge Discovery [2012].
[5] Quick, D., & Choo, K. R.. Dropbox analysis: Data remnants on user machines. Digital
Investigation, 10(1), pp 3-18 [2013].
[6] Quick, D., & Choo, K. R. Google Drive: Forensic analysis of data remnants. Journal of
Network and Computer Applications, 40, pp 179-193 [2014].

Rose Shumba
[7] Quick, D., Martini, B., & Choo, K. R. Forensic Collection of Cloud Storage Data. Cloud
Storage Forensics, pp 153-174 [2014].
[8] Chung, H., Park, J., Lee, S., & Kang, C.. Digital forensic investigation of cloud storage
services. Digital Investigation, 9(2), pp 81-95 [2012].
[9] Malik, R., Shashidhar, N., & Chen, L.. Analysis of Evidence in Cloud Storage Client
Applications on the Windows Platform. In the Proceedings of the Int'l Conf. Security and
Management [2015].
[10]Epifani, M.. Cloud Storage Forensics. Paper presented at SANS European Digital Forensics
Summit, Prague [2013].
[11]Long, C., & Qing, Z.. Forensic Analysis to China’s Cloud Storage Services. International
Journal of Machine Learning and Computing, 5(6), pp 467-470 [2015].

Rose Shumba
TABLE 1: Summary of Experiments
Resear
cher
Storage
Service
Device used for access
PC Mobile device
Platform Browser
Win
7
Win
10
Win
XP
Win
8
Mac
OS
X
Firef
ox
IE Chro
me
Safar
i
Andro
id
iOS Windo
ws
Chung Amazon
s3
    
Dropbox      
Evernote      
Google
docs
  
Quick Dropbox     
SkyDrive  
Google
Drive
    
Epifani Dropbox    
Google
Drive
   
SkyDrive    
iCloud   
Long Baidu 
360 
Malik Copy  
ownClou
d
 
Dropbox  

Rose Shumba
TABLE 2: Summary of Activities for Google Drive
Activity Google drive (Choo) Google Drive (Ephani)
Installation  Client software prefetch file and link files
 Created registry keys.
 Program files folder for installtion
 Sync_config.db - user email used to access
the Google Drive account;
 Installs in Program File folder
 Client configuration in user profile
 New registry keys created
 Registry –installed version
 Link file and prefetch files for the
client software
 sync_config.db –client version,
sync root path, user email
 snapshop.db –file names, created,
modified, URL, size, shared
Uploading
(sample
documents)
 Thumbcache.db -thumbnails for any uploaded
pictures
 Snapshot.db- URL and Resource-ID identifier.
Downloading
(client
software and
test data file)
 Prefetch files and link files for filenames of
client executable and test data files.
 RecentDocs key in NTUSER,dat for recently
used documents.
 Network Traffic analysis; IP addresses
registered to Google and URL were observed
 URL, from which client software from cookies,
history, icons, pagefioe.sys, unallocated,
Temp Internet files.
Deletion  Link file references remaining after deletion
 Prefetch files for client and uploaded files
remained.
 Deleted information on Google Drive and test
files in NTUSER.dat registry files.
 Deleted files in $Recycle.Bin folder with the
SID of user
 snapshot.db remained -filename
Accessing:
storage
through a
browser,
 Username –Web browser, history, cookies
 URL for client software –Cookie files, Web
history,
 Downloaded files –web history
 IP addresses and registered owners (Network
 Sync_log.log –sync sessions, files
created, file saved, file deleted,
deleted files, version history,
recent activities,
 RAM Analysis (username and
password in clear text)

Rose Shumba
traffic).
 RAM analysis -full text of sample files,
username in pagefile.sys file, and filenames
for the test files
Access
through the
client
 Login sessions
 URL for security certificate
 RAM analysis – session
information, user email, version
number, snapshot.db,
sync_config.db path, local sync
folder path
Uninstall  Client config folder is removed
 Prefetch and link files not deleted,

Rose Shumba
TABLE 3: Summary of Activities for Dropbox
Activity Dropbox (Quick, Choo) Dropbox (ephani)
Installation  Installs in the user profile
 Filecache.dbx - server path, local filename,
local creation time, local modified time, local
size
 Host.db –path for Dropbbox file storage
 Installed user profile
 New registry keys created
 Registry –installed version, install
location
 Prefetch and link file for the client
software
 Host.db – local folder used to
sync the account
 Filecache.dbx –server path, local
filename, local creation time, local
modified time, local size
Upload  Dropbox URL-from file listing.
 References to Dropbbox URL, software files
and folders, sample files, and test data files.
 NTUSER.dat provided a list of Dropbox and
sample files.
 Prefetch files for Dropbox executable,
Dropbox sample files, and test data files.
 Link files for filenames and folder names for
the Dropbox executable, Dropbox sample
files and sample files.
 Thumbnails for the Dropbox sample pictures,
 RAM Analysis –password in clear text,
sample filename references,
Download  Access through browser- References to
filenames.
 Website information located in the Cookie
files, Web history, FavIcons
 Filenames for downloaded files from Web
history of the browsers.
 Link files and prefetch files for Dropbox
executable, Dropbox sample files
 Thumbnails for the Dropbox sample pictures,
Deletion/erase  Deleted Dropbox information –NTUSER.dat  snapshot.db remained -filename

Rose Shumba
registry (filesnames, URL references)
 Cleaning removes all file references in the
browsers, but information remained in Google
Chrome FavIcon for Dropbox use
Accessing:
drive through
a browser,
 synched devices, timestamp of the last
activity, the IP address for the last
connection, and the version, deleted files,
timeline of previous events,
 Browser access to account: filename
references for accessed files and Dropbox
sample files, prefetch files for Dropbox
executable, Dropbox sample files, Dropbox
website information, network traffic analysis -
a session with an IP in the Range, digital
certificates,
 deleted files,
 devices connected to the account
 for every file version history
 last browser sessions
 RAM Analysis- client access (user
email, display name,
filecache.dbx path, server time,
file list, deleted file)
 RAM- browser access (login
email, login password)
Access
through the
client
 prefetch files for Dropbox executable,
Dropbox sample files,
 References to sample files in the browser
history.
 RAM analysis Session
information (User email, version
number, snapshot.db,
sync_config.db path, local sync
folder path
Uninstall  Only Dropbox.exe marked deleted, others
remained.
 Synch folder and file contents remained on
hard drive.
 All remnants were unaffected.
 Client config folder is removed
 Prefetch not deleted
 Local copy of the file not deleted
 Can recover registry keys about
recent files, link files, browser
history and cach, thumbnails,

Client Forensics: An Assessment of Existing Research And Future Directions

More Related Content

What's hot (20)

Similar to Client Forensics: An Assessment of Existing Research And Future Directions (20)

Recently uploaded (20)

Client Forensics: An Assessment of Existing Research And Future Directions