CYBEX the cybersecurity information exchange framework
CYBEX the cybersecurity information exchange framework
∗ †
Anthony Rutkowski Youki Kadobayashi Inette Furey
Yaana Technologies, USA NAIST, Japan DHS, USA
[email protected] [email protected] [email protected]
‡
Damir Rajnovic Robert Martin Takeshi Takahashi
FIRST, USA MITRE, USA NICT, Japan
[email protected] [email protected] takeshi [email protected]
This article is an editorial note submitted to CCR. It has NOT been peer reviewed.
The authors take full responsibility for this article’s technical content. Comments can be posted through CCR Online.
ACM SIGCOMM Computer Communication Review 59 Volume 40, Number 5, October 2010
Table 1: CYBEX family specifications
CYBEX family specifications
Functional blocks
imported specifications newly built specifications
Information Description block CPE, CCE, CVE, CWE, CAPEC, MAEC, CVSS, X.dexf, X.pfoc
CWSS, OVAL, XCCDF, ARF, IODEF, CEE,
TS102232, TS102667, TS23.271, RFC3924, EDRM,
Information Discovery block X.cybex.1, X.cybex-disc
Information Query block X.chirp
Information Assurance block EVCERT, TS102042 V2.0 X.eaa
Information Transport block TS102232-1 X.cybex-tp, X.cybex-beep
vspace-2mm
Software,
2. SCOPE OF CYBEX Systems,Services,
CYBEX focuses on cybersecurity information exchange Networks
ACM SIGCOMM Computer Communication Review 60 Volume 40, Number 5, October 2010
CYBEX to include observable attributes. MAEC provides a language
and format for characterizing the behaviors and actions of
Information Description block
malware with two core components consisting of enumerated
elements (vocabulary) and schema (grammar).
Information Discovery block
The Countermeasure Knowledge Base accumulates infor-
Information Query block
mation on countermeasures that corresponds to cyber risks.
To describe information in the knowledge base, CYBEX in-
Information Assurance block troduces the Common Vulnerability Scoring System (CVSS)
[9, 13], Common Weakness Scoring System (CWSS) [12],
Information Transport block Open Vulnerability and Assessment Language (OVAL) [9,
13], and eXtensible Configuration Checklist Description For-
mat (XCCDF) [9, 13]. CVSS provides for an open frame-
Figure 3: Five functional blocks of CYBEX work for communicating the characteristics and impacts of
IT vulnerabilities, while CWSS provides that for software
weaknesses. OVAL provides a language used to encode sys-
tem details and an assortment of content repositories held
seamlessly with available products, making CYBEX more throughout the community, and XCCDF provides a lan-
practical and deployable. guage for writing security checklists, benchmarks, and re-
Each of the functional blocks are elaborated on in the lated kinds of documents.
following subsections. The Product & Service Knowledge Base accumulates in-
formation on products and services. To describe informa-
3.1 Information Description Block tion in this knowledge base, CYBEX introduces Common
This functional block structures cybersecurity informa- Platform Enumeration (CPE) [9, 13] and Common Config-
tion for exchange purposes and provides the formats and uration Enumeration (CCE) [9, 13]. CPE provides a struc-
languages to describe it. These formats and languages are tured naming scheme for information technology systems,
depicted through the introduction of 18 existing specifica- platforms, and packages, while CCE provides unique iden-
tions and three newly created ones. tifiers to system configuration issues to facilitate fast and
From the viewpoint of the cybersecurity operational in- accurate correlation of configuration data across multiple in-
formation ontology that is modified from the one in [11] to formation sources and tools. Note that knowledge on cyber
accommodate forensic aspects, these specifications are clas- risks and countermeasures are often linked to specific prod-
sified as shown in Figure 4. The following subsections pro- ucts and services. For instance, a CVE is linked to CPE
vide the details of the introduced specifications following the identifiers and CVSS scores in NVD. Hence the Product &
operation domains defined by the ontology. Service Knowledge Base is linked to Cyber Risk Knowledge
Base and Countermeasure Knowledge Base as is shown in
3.1.1 Knowledge Accumulation Domain Figure 4.
Knowledge Accumulation domain is an operation domain
that accumulates knowledge on cybersecurity, which will be
3.1.2 IT Asset Management Domain
then shared and reused by other organizations. The Na- IT Asset Management domain is an operation domain
tional Vulnerability Database [10], for instance, is providing that administrates and protects IT assets of user organi-
practical facilitation for such operations. The information zations. The necessary information for this operation is
required for this operation is stored in either of the three stored in the User Resource Database and Provider Resource
knowledge bases: Product & Service, Cyber Risk or Coun- Database. To describe information in the User Resource
termeasure. Database, CYBEX introduces the Assessment Result For-
The Cyber Risk Knowledge Base accumulates informa- mat (ARF) [13], which provides a standardized IT asset as-
tion on cyber risks including that on vulnerabilities and sessment result format that facilitates the exchange of such
threats. To describe information in the knowledge base, results among systems.
CYBEX introduces Common Vulnerabilities and Exposures
(CVE) [9, 13], Common Weakness Enumeration (CWE) [9, 3.1.3 Incident Handling Domain
13], Common Attack Pattern Enumeration and Classifica- Incident Handling domain is an operation domain that
tion (CAPEC) [9, 13], and Malware Attribute Enumeration monitors and responds to cyber-incidents. The necessary in-
and Characterization (MAEC) [9, 13]. CVE provides unique formation for this operation is stored in the Incident Database
identifiers for publicly known vulnerabilities in commercial and Warning Database. To describe information in the Inci-
and open source software to facilitate rapid and accurate dent Database, CYBEX introduces the Incident Object De-
correlation of vulnerability data across multiple information scription Exchange Format (IODEF) [6], X.pfoc, and Com-
sources and tools. CWE is an XML/XSD-based specifica- mon Event Expression (CEE) [9, 13]. IODEF defines a
tion that provides unique identifiers for the weaknesses in data representation that provides a framework for exchange
software code, design, architecture, or implementation as of information about computer security incidents. X.pfoc
well as a rich body of knowledge about the cause, impact, (Phishing, Fraud, and Other Crimeware Exchange Format)
and mitigations of these weaknesses to include code exam- extends IODEF to support the reporting of phishing, fraud,
ples. CAPEC is an XML/XSD-based specification that pro- and other types of electronic crime. The extensions also
vides unique identifiers for the patterns of attack against support exchange of information about widespread spam
software as well as a rich body of knowledge about the at- incidents. CEE defines a common language and syntax for
tack steps, impact, and mitigations of these attack patterns expressing how events are described, logged, and exchanged.
ACM SIGCOMM Computer Communication Review 61 Volume 40, Number 5, October 2010
Provider Resource Product & Service KB
DB
IT Asset IT Infrastructure
CPE CCE
Management Provider
Domain User Resource DB
Researcher
ARF
Administrator
Incident DB Countermeasure KB
IODEF CEE Knowledge
Incident Response Team CWSS CVSS Accumulation
X.pfoc
Handling Domain
Registrar
Domain OVAL XCCDF
Warning DB
Coordinator
Cyber Risk KB
Evidence DB
TS102232 TS102657 Product
Forensics CVE CWE
Domain RFC3924 TS23.271 & Service
Developer
Inspector CAPEC MAEC
X.dexf EDRM
necessary information for this operation is stored in the Ev- ITU-T|ITU-R ISO
idence Database. To describe information in the database,
CYBEX introduces six forensics specifications: ETSI TS102232 1 3
.. .
[2], ETSI TS102657 [3], ETSI TS23.271 [1], RFC3924 [5], 0 2 4 48 = cybersecurity
ACM SIGCOMM Computer Communication Review 62 Volume 40, Number 5, October 2010
in X.cybex-tp. This describes the overview of transport pro-
tocols for cybersecurity information exchange. Based on the
general overview, protocol specific features are described in
RDF search the X.cybex-beep draft recommendation, which describes a
engine A transport protocol based on BEEP. Albeit other protocols
Search
can be used for this transport, currently only the BEEP
protocols are being investigated. Other candidate proto-
cols, such as SOAP, exist but no draft recommendation for
Cybersecurity List of RDF search
Entity • Identifiers engine B
such protocols have been presented yet. From the viewpoint
• capability of forensics, ETSI TS102232-1 is also introduced here. This
information
provides assurance of forensics information delivery to law
enforcement and security authorities.
RDF search
engine C
4. USE CASES
Cybersecurity information source Target range of search CYBEX provides the framework for exchanging cyberse-
curity information between cybersecurity entities. The us-
Figure 6: RDF-based discovery age of the standard is up to users. Nevertheless, to demon-
strate the usability of CYBEX, this section describes two
use cases of CYBEX.
an RDF search engine, which has its own list of indices to
the assorted cybersecurity information in the network. Note
Manipulate Transport the Receives any
that the search ranges of each RDF search engine are dif- Discover
cybersecurity information from updates of the
cybersecurity
ferent. Then the search engine replies to the cybersecurity information of the the entity to the information
entity
entity with the list of identities and capability information discovered entity user
ACM SIGCOMM Computer Communication Review 63 Volume 40, Number 5, October 2010
the entity via SSL. CERT A then receives EVSSL from the usage, CYBEX will not be able to provide its true value or
entity, with which it can ensure that the entity is CERT B contribute to cybersecurity. In order to advance cybersecu-
(the Information Assurance block). CERT A thus sends the rity, the effectiveness of CYBEX needs to be globally and
incident information following the IODEF format to CERT widely recognized.
B, which sends back another IODEF message to report the
completion of implementing countermeasures later (Infor- 7. ADDITIONAL AUTHORS
mation Description block). The procedure is depicted in
Additional authors: Craig Schultz (Multimedia Architec-
Figure 8.
tures, email: [email protected]) and Gavin Reid (Cisco,
email: [email protected]) and Gregg Schudel (Cisco, email:
[email protected]) and Mike Hird (BIS, email: mike.
CERT A RDF Search Engine CERT B [email protected]) and Stephen Adegbite (FIRST, email:
[email protected]).
Discovery
ACM SIGCOMM Computer Communication Review 64 Volume 40, Number 5, October 2010