# Online Acquisition of Digital Forensic Evidence

## Full Text

This Markdown is an automated text extraction for search and reading tools. The [hosted PDF](OnlineAcquisitionOfDigitalForensicEvidence.pdf) is the authoritative publication source.

Source PDF: [OnlineAcquisitionOfDigitalForensicEvidence.pdf](OnlineAcquisitionOfDigitalForensicEvidence.pdf)
Source SHA-256: `8f67c110e51f90c3b0ae08e7c789aa41e75fa578e1246c6425622f0ff88411e3`
Generated with `pdftotext -nopgbrk`; formatting and reading order may differ from the PDF.

---

Online Acquisition of Digital Forensic Evidence
Mark Scanlon and Mohand-Tahar Kechadi
UCD Centre for Cybercrime Investigation,
School of Computer Science and Informatics,
University College Dublin, Belfield, Dublin 4, Ireland.
{mark.scanlon,tahar.kechadi}@ucd.ie

Abstract. Providing the ability to any law enforcement officer to remotely transfer an image from any suspect computer directly to a forensic
laboratory for analysis, can only help to greatly reduce the time wasted
by forensic investigators in conducting on-site collection of computer
equipment. RAFT (Remote Acquisition Forensic Tool) is a system designed to facilitate forensic investigators by remotely gathering digital
evidence. This is achieved through the implementation of a secure, verifiable client/server imaging architecture. The RAFT system is designed
to be relatively easy to use, requiring minimal technical knowledge on
behalf of the user. One of the key focuses of RAFT is to ensure that
the evidence it gathers remotely is court admissible. This is achieved by
ensuring that the image taken using RAFT is verified to be identical to
the original evidence on a suspect computer.
Key words: Digital Forensics, Evidence, Remote, Hard Drive Acquisition, Imaging, Internet, Verification.

1 Introduction
Current trends in technology are putting computers with high-bandwidth Internet connections into the hands of regular criminals. As this phenomenon continues, an increasing number of traditional crimes are being aided by computers,
e.g., fraud, identity theft, terrorism, etc. As a result, digital forensic investigators are becoming overwhelmed with the number of cases they have to deal with.
Traditional digital forensic investigations commence with the investigators leaving their laboratory to visit the crime scene, where they collect all the relevant
evidence, and bring it back to the forensic laboratory for secure storage and
analysis. This evidence may then lay untouched for extended periods while the
investigating team deals with the backlog of cases.
In this paper, we introduce a solution to reduce the time taken to acquire
the necessary evidence. We propose RAFT (Remote Acquisition Forensic Tool),
a remote forensic hard drive imaging tool, that is designed to boot off a Linux
Live CD or USB memory stick. A brief overview of the RAFT scenario is depicted in Fig. 1. The suspect computer is booted using a customised Linux Live
distribution and any hard drives or removeable media connected to the computer are able to be securely imaged over an internet connection directly to the

2

Mark Scanlon et al.
Suspect Computer
booted using RAFT

RAFT Server in
Forensic Lab

Secure Internet Connection

Fig. 1. Fundamental design of the RAFT system.

RAFT Server. This system is designed to equip any law enforcement or investigating officers with the ability to easily perform digital evidence acquisition,
which would traditionally require the expertise of an on-site forensic investigator. One key objective of the RAFT system is to ensure that the evidence it
gathers is court admissible. This is achieved by ensuring that the image taken
using RAFT is forensically verified to be identical to the original evidence.

2 System Architecture
The RAFT System is based on a client/server architecture as can be seen in
Fig. 2. The client side of the RAFT System is designed to be easy to use and
to require minimal training in booting from the CD or USB memory stick. The
RAFT Server is a multi-threaded server which can accept multiple connections
simultaneously.
2.1 RAFT Client
The RAFT Client is installed on a customised lightweight copy of the Ubuntu
Live Linux distribution [11]. Ubuntu was chosen for a number of reasons:
1. The standard Ubuntu install disk comes packaged with a live linux distribution. This live disk is bootable on any modern computer, regardless of
manufacturer or operating system, e.g., Windows, *nix or MAC OS.
2. The compatability of the live disk to read numerous different drive formats,
e.g., FAT, FAT16, FAT32, ext, ext2, ext3, HFS, HFS+, etc.
3. The ability for the live distribution to be fully customised removing any
unnecessary software, while having the ability to easily include the RAFT
Client and associated software (such as dcfldd [2] and sshfs [10]).
4. The ability to configure the auto mounting script to automatically mount
all attached drives and storage as read-only.

Online Acquisition of Digital Forensic Evidence

3

RAFT Server running on
server in forensic lab

...

...

Secure File Transfer over
Internet Connection

Original Hard
Drive

RAFT Client running on
suspect PC

Hard Drive
Chunks
Verified Hard
Drive Image

Fig. 2. Overview of RAFT imaging architecture. The left side of the figure represents
the suspect computer, while the right represents the server-side of the system in a
forensics laboratory.

When the Ubuntu Live CD is booted on the suspect computer, all the attached hard drives and removeable drives, e.g., USB memory sticks, external
hard drives, cameras, CDs etc., which are currently connected are automatically
mounted as read-only. The user is then prompted within the graphical user interface whether s/he would like to take an image of a specific drive or partition, or
if s/he would like all the attached media to be imaged automatically. To ensure
forensic verifiability, a hash is taken of the entire drive before it is imaged. This
initial hash is taken to verify the untouched drive to the evidence ultimately
collected from it.
2.2 RAFT Server
The RAFT Server is a multi-threaded system. When the server is running, it
listens for a connection on any of its preconfigured ports. When a new client is
connected, it creates a new space on the server where it stores all relevant files.
Each new drive or volume imaged from the suspect computer is then stored
within that space. Each chunk, when successfully transferred, is hashed serverside and compared against the original hash. If these hashes differ, a failure
notification is sent to the RAFT Client which will result in that particular chunk
being re-transmitted.
Upon the successful transmission of all the chunks belonging to a particular
drive, they are recompiled back into a single file, hashed and verified against
the original hash value taken by the RAFT Client before the imaging process
commenced. When these two hashes match, a “successfully completed” transfer
notification is then sent to the RAFT Client.

4

Mark Scanlon et al.
RAFT Client
Drive Detection
and Auto-mount
Module

GUI Module

Detect all hard
drives, partitions,
and all attached
removable storage
devices i.e. CD/DVD
drives, memory keys,
cameras etc.

The GUI presents a
list of all the
discovered storage
devices to the user
and gives him an
option for which
devices to image.

Mount all attached
storage devices as
read-only.

The user also has an
“image all” option.
This will
automatically image
all the attached
storage devices
without any further
interaction.

RAFT Server

Hashing and
Imaging Module
Before each drive is
imaged, a SHA-512
hash is taken and
sent to the server.
The first chunk is
transmitted to the
server and is hashed
during transmission
using dcfldd. This
hash is then sent to
the server.
The client waits for a
notification from the
server if the transfer
succeeded.

SSH Connection for
Client/Server
Communication

SFTP Connections for
Evidence Transmission

Forensic
Integrity
Module

Recombination
and Storage

Each chunk is
received and a
SHA-512 hash is
calculated. This hash
is then compared to
the original hash.

Recombines all the
chunks received, in
the correct order, to
form the original
image.

If the hashes match,
an acknowledgement
is sent to the RAFT
Client. If the hashes
differ, a failure
notification is sent to
the client which
requests the retransmission of that
chunk.

A final SHA-512 hash
is taken of the
compiled image and
verified against the
first hash.
If it matches, the
image acquisition
from the suspect
computer has been
successful.

Fig. 3. Overview of the steps involved, both client and server side, in verifiable image
acquisition using RAFT.

2.3 Forensic Integrity
Due to the sensitive nature of digital evidence collection, it is imperative that
the data collected by any forensic tool is completely verifiable and identical
to the original source. This integrity is insured in the RAFT system by the
implementation of regular hash checking on the data being transferred using
SHA-512, a 512-bit secure hashing algorithm. Once the RAFT Client is booted
and a drive is selected for imaging, the first step is to calculate a hash value
for the original drive. During the imaging process, the integrity of each of the
chunks being transfered is maintained due to a SHA-512 hash being computed as
the chunk is being transmitted. Server-side, once the transmission is completed,
a SHA-512 hash is taken on the chunk and verified against the original. If these
hashes do not match, i.e., the integrity of that chunk has been compromised in
transmission, a failure notification is sent to the client, which queues that chunk
up again for transmission.

3 Advantages
– Compatibility – One obvious advantage of using the RAFT system is that it
is irrelevant what configuration the suspect PC has, i.e., RAFT is compatible
with whatever interface or formatting the suspect hard drive or media might
have. Take netbooks as an example: netbooks come in many various storage
configurations, even within the same brand. Some netbooks use regular 21/2"
IDE or SATA laptop hard drives whereas some use flash storage. These flash
drives can be soldered directly to the motherboard, connected via a regular
IDE or SATA connection or connected via a mini-PCI/mini-PCIe connection.
RAFT has no limitation on which configuration is used, the system will automatically mount and securely image any configuration.

Online Acquisition of Digital Forensic Evidence

5

– Cost – The cost involved in running the RAFT system is mainly on the server
side. The RAFT Server would need to run on a high-end computer with a
very high-speed internet connection. It would also be required to have a large
amount of available storage, be it local storage or a connected NAS (network
attached storage). However, once the initial outlay is spent in setting up the
RAFT Server, the cost for using and re-producing the RAFT Client is minimal.
For example, in a law enforcement scenario, the customised RAFT ubuntu
image can be burnt to CD or a bootable USB key can be created as many
times as required, e.g., one of each per police station.
– Automated Acquisition – This feature of the RAFT system results in users
requiring little technological knowledge to operate the client side of the system.
Due to the ease of use of the system, it will ultimately result in forensic
acquisition being possible in more places at once, e.g., in the law enforcement
scenario outlined above, each police station would have the capability to image
a computer without the need to have a digital forensic specialist.
– Speed – While each individual image acquisition can take some time, multiple
acquisitions can take place simultaneously. This results in an overall decrease
in the time taken for multiple computers to be imaged at once to the same
server.
The combined advantage of the above points results in the forensic investigator
being able to spend more time in the laboratory analysing the evidence collected.
Using RAFT in combination with more intelligent forensic analysis tools, e.g.,
a distributed digital forensic system [3], [6], [7], the investigator will be better
armed to deal with an increasing amount of casework.

4 Potential Limitations
While the RAFT system has several advantages, such as those outlined above,
there are also some potential limitations:
– Firewall – The RAFT Client has to have the ability to communicate to the
server, for the transmission of the evidence. One obvious potential limitation
of the system is that a hardware firewall may be filtering the suspect computer’s internet connection, e.g., blocking specific port ranges, etc. This could
potentially render the RAFT Client inoperable. One solution to this is to employ the use of a USB mobile broadband connection, connected to the suspect
computer. Current 3G wireless broadband networks are capable of upstream
speeds of up to 10 Mbps, with plans for 3G LTE (Long-Term Evolution) to increase the upstream speeds to over 50 Mbps [1]. These potential upload speed
are set to improve even further when 4G mobile broadband networks become
mainstream in the coming years. 4G networks will be capable of upload speeds
of over 100 Mbps [4].
– Transfer Speed – The time taken to take an image of a hard drive over the
Internet will take longer than the time required if the investigator had physical

6

Mark Scanlon et al.

access to the device and the imaging was conducted in a laboratory. RAFT can
improve on this time required for traditional hard drive image acquisition if the
time wasted by the investigation in travelling, transportation and storage of
the suspect computer is taken into consideration. While high-speed broadband
internet access is becoming more and more common place on both residential
and commercial levels, it would be unrealistic to assume that every suspect
computer would have an internet connection with a favourable upload speed,
e.g., many asymmetric broadband connections are weighted towards download
speed resulting in significantly slower upstream speeds. This limitation could
again be overcome through the use of a mobile internet connection.
– Non-functional/no CD Drive – It is possible that some suspect computers have
a non-functional CD drive. For that matter, some modern computing equipment do not feature a CD drive, e.g., most netbooks and small laptops. This
issue can be overcome as the customised Ubuntu operating system containing
RAFT can also be configured to boot of a USB memory key.
– Live System – Forensic investigators are increasingly concerned with the analysis of live systems, e.g., collection of evidence of current processes, memory
and other state information. In its current from, the RAFT system is unable
to collect evidence from a live system. However, an extended version of the
RAFT client could be provided to run on a live system. The downside of
executing RAFT on a live system is that there will be an unavoidable, yet
predictable, change of state of the live system.
– Boot Passwords – Should the suspect PC have a CMOS boot password, before
the user has the opportunity to boot up the RAFT Client, he must insert this
password. In the quite likely event that this password is unknown, it will
be necessary to reference documentation to retrieve the BIOS manufacturer’s
backdoor CMOS password. A sample list of common BIOS manufacturer’s
and their associated backdoor passwords can are given in table 1 [12]:
Table 1. Backdoor BIOS passwords for common motherboard manufacturers
Manufacturer Commonly Used Passwords
AWARD
01322222, 589589, 589721, 595595, 598598 , ALFAROME,
ALLY, ALLy, aLLY, aLLy, aPAf, award, AWARD PW, AWARD
SW, AW ARD?SW , AW ARD P W , AW ARD SW , AWKWARD, awkward, BIOSTAR, CONCAT, CONDO, Condo,
condo, d8on, djonet, HLT, J256, J262, j262, j322, j332, J64,
KDD, LKWPETER, Lkwpeter, PINT, pint, SER, SKY F OX,
SYXZ, syxz, TTPTHA, ZAAAADA, ZAAADA, ZBAAACA,
ZJAAADC
AMI
AMI, AAAMMMIII, BIOS, PASSWORD, HEWITT RAND,
AM I?SW , AM I SW , LKWPETER, A.M.I., CONDO
PHOENIX
BIOS, CMOS, phoenix, PHOENIX, Phoenix

Online Acquisition of Digital Forensic Evidence

7

While such a list may seem combersome, a database of computers and motherboards and their corresponding backdoor passwords could be created to
quickly access the correct password. If the password remains undocumented,
a reset of the CMOS could be performed by removing the CMOS battery from
the motherboard. After a short period of time, e.g., less than ten minutes, the
BIOS will be reset to its factory state, with no boot password. Some motherboards also incorporate a jumper which, when removed, enables the user to
bypass the CMOS password. Passwords further in the regular boot process
of the suspect system, e.g., an operating system login password, will have no
effect on the operation of the RAFT system as it will be the customised Live
Ubuntu operating system which is booted immediately after the BIOS.

5 Results

Time Taken in seconds

To evaluate the performance of the RAFT system, numerous “real-world” scenarios were tested. For the purpose of this paper, we will discuss two of these
scenarios. During the testing of the RAFT system, the performance of the imaging process tended to be linear. As a result, all of the results discussed below
have been averaged to reflect the performance for each gigabyte of digital evidence collected. The “dcfldd” tool used in the RAFT system has the ability
to compute the hash values at the same time as transmitting the chunk. The
four values displayed in Figures 4 and 5 show the impact of the various hashing
options on the overall performance.

100
90
80
70
60
50
40
30
20
10
0
No Hashing

SHA-256

SHA-512

SHA-256 &
SHA-512

Fig. 4. Comparison of imaging times per gigabyte from a high-speed internet connection with a 83.26Mbps upload speed as tested using speedtest.net.

8

Mark Scanlon et al.

Time Taken in seconds

The first scenario concerns imaging a suspect computer with a very highspeed internet connection with a 87.62Mbps downlink and a 83.26Mbps upload
streams (connection speed tested using Speedtest [9]). The suspect computer in
this scenario was a Dell Optiplex 745 with a 2.66Ghz Intel Core 2 Duo processor,
2GB 667Mhz memory and a 250GB 31/2" 7200rpm hard drive. As can be seen
in Fig. 4, the average time required per gigabyte was 92 seconds. The variance
in the experiments conducted was ± 1 second or 1.09%.

1214
1213
1212
1211
1210
1209
1208
1207
1206
1205
No Hashing

SHA-256

SHA-512

SHA-256 &
SHA-512

Fig. 5. Comparison of imaging per gigabyte times from a residential broadband connection with a 8.14Mbps upload speed as tested using speedtest.net.

The second scenario involved imaging a suspect computer from a residential broadband connection, with obviously low bandwidth speeds. The suspect
computer in this scenario was a Dell XPS M1330 with a 2.5Ghz Intel Core 2
Duo processor, 4GB 667Mhz memory and a 320GB 21/2" 7200rpm hard drive.
The broadband connection had a download speed of 23.82Mbps and an upload
speed of 8.14Mbps (connection speed tested using Speedtest [9]). The result from
these tests was that the average time to acquire a 320GB hard drive image was
approximately 20 minutes per gigabyte, as can be seen in Fig. 5. The variance
in the experiments conducted was ± 8 seconds or 0.0066%.
One requirement of the performance evaluation of the RAFT system was to
quantify the overhead added through the secure hashing of each chunk. It was
found that the cost for the hashing of each chunk averaged at 5.3 seconds per
gigabyte (or a 0.41% increase in the time taken) as can be seen in Fig. 5.
The time taken for the server to verify each of the hard drive chunks received
is approximately 20 seconds per gigabyte using the SHA-256 hashing algorithm,

Online Acquisition of Digital Forensic Evidence

9

60

Time Taken (seconds)

50
40
MD5
30

SHA-1
SHA-256

20

SHA-512

10
0
1GB

2GB

4GB

8GB

16GB

32GB

Fig. 6. Server side comparison of selected common hashing algorithms on various file
sizes (normalised to a per-gigabyte value).

as can be seen in Fig. 6. This figure also displayed the server-side hashing times
as a comparison of three other common hashing algorithms. One can notice that
there is no additional overhead involved in choosing to use the 256-bit algorithm
(SHA-256) as opposed to the 128-bit algorithm (MD5) and the 160-bit algorithm
(SHA-1). However, there is a substantial cost of over double the computational
time in using the more collision resistant 512-bit algorithm (SHA-512). The
extra time required to compute the SHA-512 hash will only impact the overall
imaging time twice; once on the RAFT Client before the imaging process begins,
and again server-side when the image is completed.

6 Conclusion and Future Work
While the time taken to image a suspect computer over the Internet is substantially longer than the time taken using traditional forensic methods (with direct
physical access to the hard drive [5]) the traditional approach does not factor
in the time wasted by forensic professionals in the collection of this evidence.
Using RAFT could give forensic investigators the power to remotely conduct
investigations in more places at once.
One significant plus of using the RAFT system is that it is extremely cost
effective to distribute the client side of the system over many locations. In the law
enforcement scenario, this could result in every police station having a copy of
the RAFT Client. This would result in granting regular police officers the ability
to quickly image a suspect computer, after receiving any necessary warrents.

10

Mark Scanlon et al.

The RAFT System could be improved upon by giving optional total control of the RAFT Client to a remote investigator (after a suspect computer is
booted). The investigator could remotely browse files on the suspect PC without
the requirement to take an entire hard drive image. This would enable the investigator to determine if the suspect computer is relevant to the case and could
help focus the investigation quickly on the computer(s) relevant to the crime being investigated. The imaging process would also be streamlined, focusing onto
the necessary hard drives or partitions in the suspect computer.
A version of the RAFT Client could be created with the specific purpose of
imaging a live system. This would have the ability to collect additional evidence
from a live system, e.g., evidence located in memory, running process information
and other state information. While executing any program on a live system will
alter its state, this alteration would be predictable and should not interfere with
the original evidence.
The system could also be improved upon by eliminating unnecessary transfer
size, thus improving on the time required to collect all necessary evidence, e.g.,
through the employment of a lossless forensically sound compression algorithm.

References
1. Dahlman, E., Ekström, H., Furuskär, A., Jading, Y., Karlsson, J., Lundevall, M.,
Parkvall, S.: The 3G Long-Term Evolution – Radio Interface Concepts and Performance Evaluation. In: 63rd IEEE Vehicular Technology Conference, Melbourne,
Australia (May 2006)
2. dcfldd (Department of Defence Computer Lab Dataset Definition), April 2009,
http://dcfldd.sourceforge.net
3. Gao, Y., Richard III, G.G., Roussev, V.: Bluepipe: A Scalable Architecture for Onthe-Spot Digital Forensics. In: International Journal of Digital Evidence, vol. 3 no.
1 (2004)
4. Govil, J., Govil, J., An Empirical Feasibility Study of 4G’s Key Technologies.
In: Proceedings IEEE International Conference on Electro/Information Technology
(EIT), pp. 267–270 (2008)
5. Ray, I., Shenoi, S., Advantages in Digital Forensics IV, Chapter 26: Time Analysis
of Hard Drive Imaging Tools, pp. 340 – 343 (2008)
6. Richard III, G.G., Roussev, V., Next-Generation Digital Forensics. Communications
of the ACM, vol. 49 no. 2, pp. 76 – 80 (2006)
7. Roussev, V., Richard III, G.G., Breaking the performance wall: the case for distributed digital forensics. In: Proceedings of the 2004 Digital Forensics Research
Workshop, Baltimore, Maryland, USA (DFRWS August 2004)
8. Sealey, P., Remote Forensics. In: Digital Investigation vol. 1 no. 4, pp. 261–265.
Elsevier (2004)
9. Speedtest Mini, Downloaded April 2009, http://www.speedtest.net/mini.php
10. SSH Filesystem, Downloaded April 2009 http://fuse.sourceforge.net/sshfs.html
11. Ubuntu Linux 8.04, November 2008, http://www.ubuntu.com
12. Wang, S-J., Measures of retaining digital evidence to prosecute computer-based
cyber-crimes. In: Computer Standards & Interfaces, vol. 29 no. 2, pp. 216-223
(February 2007)
