Summer VFRP Experience

Tool Development for a Cyber SA System
Martin Q. Zhao
October 1, 2010
VFRP when and where
 Applied for SFFP (summer faculty fellowship program)
jointly sponsored by ASEE (American Society of
Engineering Education) and AFRL
 Application submitted: December, 2009
 Accepted (through VFRP): March, 2010
 Thanks to Drs. Allen, Cozart and Digh for their help
 Worked at AFRL’s Rome Research Site for 10 weeks
(May 24 – July 30)
 Griffiss Business and Technology Park
http://www.griffissbusinesspark.com/
AFRL/RI an overview
 US Air Force Research Laboratory
Information Directorate in Rome, NY.
 AFRL/RI is the component responsible for
command, control, communication and
computers and intelligence (C4I) research
and development.
 Core Technology Competencies (CTCs):
-Information Exploitation
-Information Fusion & Understanding
-Information Management
-Advanced Computing Architectures
-Cyber Operations
-Connectivity
-Command and Control
Information Fusion
 Data fusion is a formal framework in which are
expressed the means and tools for the alliance of data
originating from different sources.
 Data fusion aims at obtaining information of greater
quality; the exact definition of 'greater quality' will
depend upon the application.
 In the context of military applications, it emphasizes
collecting and processing raw data from various
sensory sources and tracking and identifying activities
of interest, so as to enable situation awareness (SA) for
the decision maker to take appropriate actions.
Unified SA Model by Salerno et al ['05]
Dr. Endsley’s model['95] :
-Perception
-Comprehension
-Projection
Dr. Salerno also co-chaired
a Social Computing
conference for 3 times
JDL (joint director of labs) model['91, revised '98]:
Level 0: Source Preprocessing/subobject refinement
Level 1: Object refinement
Level 2: Situation refinement
Level 3: Impact Assessment
Level 4: Process Refinement
Cyber SA Virtual Terrain
The virtual terrain is
a graphical
representation of a
computer network
containing
information relevant
for a security
analysis of a computer
network, including:
-Hosts & Subnets
-Routers, sensors &
firewalls
-Physical & wireless links
-Services & exposures
-Users and accounts
-Mission & criticality scores
Sample Virtual Terrain cs.mercer.edu
Internet
xxx.xxx.xxx.xxx
Lab 100
168.15.2.1 -.21
Main Switch
Lab 204
168.15.1.1
168.15.4.1 -.21
Cobra
Raptor
Intruder
168.15.1.2
168.15.1.4
168.15.1.6
Eagle
Apache
Zeus
168.15.1.3
168.15.1.5
168.15.1.7
Faculty - 1
2nd Flr. Switch
168.15.5.1 - .8
Lab 200
168.15.6.1 - .17
168.15.3.2
Lab 306
168.15.8.1 -.21
3rd
Flr. Switch
168.15.7.2
Faculty - 2
168.15.9.1 - .4
Lab 304
168.15.10.1 - .15
Sample Mission Tree cs.mercer.edu
mission
Submission_1
App_1_1
…
Asset
Submission_n
…
App_1_m
…
Asset
Cyber SA Tracking Attack Events
(1) ICMP Ping NMAP (62.34.46.54  45.34.12.1)
(2) SCAN nmap fingerprint attempt (38.244.61.9  45.34.12.2)
(3) x86 mountd overflow (62.34.46.54  45.34.12.1)
(4) gobbles SSH overflow (62.34.46.54  45.34.12.1)
(5) SCAN cybercop os SFU12 probe (38.244.61.9  45.34.12.2)
(6) WEB-MISC windmail.exe access (38.244.61.9  45.34.12.2)
(7) ICMP Ping NMap (45.34.12.1  45.34.13.1)
(8) EXPLOIT RADIUS MSID overflow attempt (45.34.12.2  45.34.12.2)
(9) chown command attempt (62.34.46.54  45.34.12.1)
(10) MS-SQL:PROCEDURE-DUMP (45.34.12.2  45.34.12.2)
IDS
alerts
Cyber SA Attack Method Categories
Cyber SA Attack Guidance Template
SITA situation identification & threat assessment
Summer Research An Overview
 Title of the proposal:
Knowledge Representation & Reasoning for Impact/ Threat Assessment
in Cyber Situation Awareness Systems
 Objective:
Enhancing the SITA system
 Find ways to model domain knowledge
 Develop a tool for VT creation/modification
 Collaborators:
 Dr. John Salerno
 Mike Manno
 Jimmy Swistak
 Warren Geiler
Problems to Solve
•Tools need to be developed to feed SITA with data
•Amount of data is huge
A computer network can have hundreds of machines, thousands
of software applications and user accounts
Known vulnerabilities are in the thousands, and the number is
ever growing.
•XML files are used: they can contain redundant data
Harm efficiency
Hard to change anything: due to well-known anomalies
oInsertion
oDeletion
oUpdate
Conceptual Data Model
Relational Data Model-VT
S/W
H/W
Link &
Policy
Exposure
Relational Data Model-Mission
Relational Data Model-Exposure
Mission Map Editor-Requirements
• (Type of) User:
SA Operator
• System Functions:





Access data in file/DB
Display a mission tree
Modify a mission tree
Save changes to file/DB
Create a mission tree
Requirements modeling
w/ a use-case diagram
Mission Map Editor-Tree creation
1
File | New
2
3
6
File | Save
5
Assign assets
Top mission
Add more
4
Set criticality
Mission Map Editor-Architecture
XML
Mission
Map Model
VT
Model
DB
Mission Map Editor-Dynamics
Vulnerability Lookup-Overview
• What is a vulnerability?
National Vulnerability
Database (NVD) contains
CVE Vulnerabilities 43054
CPE Names
• What is an exposure?
• How is it stored in NVD?
• What is CVE?
• What is CPE?
• How are they related to
SITA?
22181
Common Vulnerabilities
and Exposures (CVE)
<entry id="CVE-2010-0278">
……
<cpe-lang:logical-test negate="false" operator="OR">
<cpe-lang:fact-ref name="cpe:/o:microsoft:windows_7"/>
<cpe-lang:fact-ref name="cpe:/o:microsoft:windows_vista"/>
……
</entry>
Common Platform
Enumeration (CPE)
<cpe-item name="cpe:/o:microsoft:windows_7">
<title xml:lang="en-US">Microsoft Windows 7</title>
……
</cpe-item>
Vulnerability Lookup-Prototype
0
Load files
C
A
CVSS Rating
B
Apps
affected
Exposure
Vulnerability Lookup-Ideal ways
Type
Vendor
Prod. Line
Product
Application : a
Alcatel
MS-DOS
Windows 98
Hardware : h
Apple
Windows
Windows 2000
O/S : o
……
Windows XP
IBM
Windows Vista
……
Windows 7
Microsoft
……
CVE Entry
CVE-2010-0278
cpe:/o:microsoft:windows_7
CVE-2010-0018
CVE-2010-0249
CVE-2010-0232
……
Future R&D
•MissionMapEditor: Thorough testing and refactoring
•VulnerabilityTracker:
Research the processes of checking/updating CVE and CPE data
feeds
Design a layered system architecture
Design and implement GUI that organizes products by category
(such as OS, apps, HW), vendor, product family, version, etc
•IDS (e.g. Snort) alerts specifics and mapping with CVE, as
well as with SITA
•VT model generation using automatic scanning data
•Cyber situation visualization
Q&A
Fall Extension Updates – Vul’Tracker
Fall Extension Updates – Vul’Tracker
Fall Extension Updates – Vul’Tracker
The data feed file download and DB loading/update functions have been tested
with
•CVE data feed files for
•2010 (two versions, one from July [15 MB] and another from December
revision [39 MB]) and
•2009 [34 MB]; and
•CPE file from July 2010 [6.8 MB].
Table 1 – Vendor Counts
by Platform Types
Year(s)
2009
2010
2009~10
Type
Table 2 – Count
of Vulnerable
Software by Year
Count of
Vulnerabilities
Vendor
Count
S/W Count
Count of Vul.
Software
Count of CPE-CVE
Mapping Entries
4,082
3,054
5,511
35,833
24,999
60,832
4,606
1,993
6,599
a:
Application
403
h:
Hardware
75
15,581
1,944
o:
OS
38
4,621
Total
-22,146
Total
Distinct
Vendor
Count
432