Network Working Group R. Shirey Request for Comments: 4949 August 2007 FYI: 36 Obsoletes: 2828 Category: Informational Internet Security Glossary, Version 2 Status of This Memo This memo provides information for the Internet community. It does not specify an Internet standard of any kind. Distribution of this memo is unlimited. Copyright Notice Copyright (C) The IETF Trust (2007). RFC Editor Note This document is both a major revision and a major expansion of the Security Glossary in RFC 2828. This revised Glossary is an extensive reference that should help the Internet community to improve the clarity of documentation and discussion in an important area of Internet technology. However, readers should be aware of the following: (1) The recommendations and some particular interpretations in definitions are those of the author, not an official IETF position. The IETF has not taken a formal position either for or against recommendations made by this Glossary, and the use of RFC 2119 language (e.g., SHOULD NOT) in the Glossary must be understood as unofficial. In other words, the usage rules, wording interpretations, and other recommendations that the Glossary offers are personal opinions of the Glossary's author. Readers must judge for themselves whether or not to follow his recommendations, based on their own knowledge combined with the reasoning presented in the Glossary. (2) The glossary is rich in the history of early network security work, but it may be somewhat incomplete in describing recent security work, which has been developing rapidly. Shirey Informational [Page 1] RFC 4949 Internet Security Glossary, Version 2 August 2007 Abstract This Glossary provides definitions, abbreviations, and explanations of terminology for information system security. The 334 pages of entries offer recommendations to improve the comprehensibility of written material that is generated in the Internet Standards Process (RFC 2026). The recommendations follow the principles that such writing should (a) use the same term or definition whenever the same concept is mentioned; (b) use terms in their plainest, dictionary sense; (c) use terms that are already well-established in open publications; and (d) avoid terms that either favor a particular vendor or favor a particular technology or mechanism over other, competing techniques that already exist or could be developed. Table of Contents 1. Introduction ....................................................3 2. Format of Entries ...............................................4 2.1. Order of Entries ...........................................4 2.2. Capitalization and Abbreviations ...........................5 2.3. Support for Automated Searching ............................5 2.4. Definition Type and Context ................................5 2.5. Explanatory Notes ..........................................6 2.6. Cross-References ...........................................6 2.7. Trademarks .................................................6 2.8. The New Punctuation ........................................6 3. Types of Entries ................................................7 3.1. Type "I": Recommended Definitions of Internet Origin .......7 3.2. Type "N": Recommended Definitions of Non-Internet Origin ...8 3.3. Type "O": Other Terms and Definitions To Be Noted ..........8 3.4. Type "D": Deprecated Terms and Definitions .................8 3.5. Definition Substitutions ...................................8 4. Definitions .....................................................9 5. Security Considerations .......................................343 6. Normative Reference ...........................................343 7. Informative References ........................................343 8. Acknowledgments ...............................................364 Shirey Informational [Page 2] RFC 4949 Internet Security Glossary, Version 2 August 2007 1. Introduction This Glossary is *not* an Internet Standard, and its recommendations represent only the opinions of its author. However, this Glossary gives reasons for its recommendations -- especially for the SHOULD NOTs -- so that readers can judge for themselves what to do. This Glossary provides an internally consistent and self-contained set of terms, abbreviations, and definitions -- supported by explanations, recommendations, and references -- for terminology that concerns information system security. The intent of this Glossary is to improve the comprehensibility of written materials that are generated in the Internet Standards Process (RFC 2026) -- i.e., RFCs, Internet-Drafts, and other items of discourse -- which are referred to here as IDOCs. A few non-security, networking terms are included to make the Glossary self-contained, but more complete glossaries of such terms are available elsewhere [A1523, F1037, R1208, R1983]. This Glossary supports the goals of the Internet Standards Process: o Clear, Concise, Easily Understood Documentation This Glossary seeks to improve comprehensibility of security- related content of IDOCs. That requires wording to be clear and understandable, and requires the set of security-related terms and definitions to be consistent and self-supporting. Also, terminology needs to be uniform across all IDOCs; i.e., the same term or definition needs to be used whenever and wherever the same concept is mentioned. Harmonization of existing IDOCs need not be done immediately, but it is desirable to correct and standardize terminology when new versions are issued in the normal course of standards development and evolution. o Technical Excellence Just as Internet Standard (STD) protocols should operate effectively, IDOCs should use terminology accurately, precisely, and unambiguously to enable standards to be implemented correctly. o Prior Implementation and Testing Just as STD protocols require demonstrated experience and stability before adoption, IDOCs need to use well-established language; and the robustness principle for protocols -- "be liberal in what you accept, and conservative in what you send" -- is also applicable to the language used in IDOCs that describe protocols. Using terms in their plainest, dictionary sense (when appropriate) helps to make them more easily understood by Shirey Informational [Page 3] RFC 4949 Internet Security Glossary, Version 2 August 2007 international readers. IDOCs need to avoid using private, newly invented terms in place of generally accepted terms from open publications. IDOCs need to avoid substituting new definitions that conflict with established ones. IDOCs need to avoid using "cute" synonyms (e.g., "Green Book"), because no matter how popular a nickname may be in one community, it is likely to cause confusion in another. However, although this Glossary strives for plain, internationally understood English language, its terms and definitions are biased toward English as used in the United States of America (U.S.). Also, with regard to terminology used by national governments and in national defense areas, the glossary addresses only U.S. usage. o Openness, Fairness, and Timeliness IDOCs need to avoid using proprietary and trademarked terms for purposes other than referring to those particular systems. IDOCs also need to avoid terms that either favor a particular vendor or favor a particular security technology or mechanism over other, competing techniques that already exist or might be developed in the future. The set of terminology used across the set of IDOCs needs to be flexible and adaptable as the state of Internet security art evolves. In support of those goals, this Glossary offers guidance by marking terms and definitions as being either endorsed or deprecated for use in IDOCs. The key words "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" are intended to be interpreted the same way as in an Internet Standard (i.e., as specified in RFC 2119 [R2119]). Other glossaries (e.g., [Raym]) list additional terms that deal with Internet security but have not been included in this Glossary because they are not appropriate for IDOCs. 2. Format of Entries Section 4 presents Glossary entries in the following manner: 2.1. Order of Entries Entries are sorted in lexicographic order, without regard to capitalization. Numeric digits are treated as preceding alphabetic characters, and special characters are treated as preceding digits. Blanks are treated as preceding non-blank characters, except that a hyphen or slash between the parts of a multiword entry (e.g., "RED/BLACK separation") is treated like a blank. Shirey Informational [Page 4] RFC 4949 Internet Security Glossary, Version 2 August 2007 If an entry has multiple definitions (e.g., "domain"), they are numbered beginning with "1", and any of those multiple definitions that are RECOMMENDED for use in IDOCs are presented before other definitions for that entry. If definitions are closely related (e.g., "threat"), they are denoted by adding letters to a number, such as "1a" and "1b". 2.2. Capitalization and Abbreviations Entries that are proper nouns are capitalized (e.g., "Data Encryption Algorithm"), as are other words derived from proper nouns (e.g., "Caesar cipher"). All other entries are not capitalized (e.g., "certification authority"). Each acronym or other abbreviation that appears in this Glossary, either as an entry or in a definition or explanation, is defined in this Glossary, except items of common English usage, such as "a.k.a.", "e.g.", "etc.", "i.e.", "vol.", "pp.", and "U.S.". 2.3. Support for Automated Searching Each entry is preceded by a dollar sign ($) and a space. This makes it possible to find the defining entry for an item "X" by searching for the character string "$ X", without stopping at other entries in which "X" is used in explanations. 2.4. Definition Type and Context Each entry is preceded by a character -- I, N, O, or D -- enclosed in parentheses, to indicate the type of definition (as is explained further in Section 3): - "I" for a RECOMMENDED term or definition of Internet origin. - "N" if RECOMMENDED but not of Internet origin. - "O" for a term or definition that is NOT recommended for use in IDOCs but is something that authors of Internet documents should know about. - "D" for a term or definition that is deprecated and SHOULD NOT be used in Internet documents. If a definition is valid only in a specific context (e.g., "baggage"), that context is shown immediately following the definition type and is enclosed by a pair of slash symbols (/). If the definition is valid only for specific parts of speech, that is shown in the same way (e.g., "archive"). Shirey Informational [Page 5] RFC 4949 Internet Security Glossary, Version 2 August 2007 2.5. Explanatory Notes Some entries have explanatory text that is introduced by one or more of the following keywords: - Deprecated Abbreviation (e.g., "AA") - Deprecated Definition (e.g., "digital certification") - Deprecated Usage (e.g., "authenticate") - Deprecated Term (e.g., "certificate authority") - Pronunciation (e.g., "*-property") - Derivation (e.g., "discretionary access control") - Tutorial (e.g., "accreditation") - Example (e.g., "back door") - Usage (e.g., "access") Explanatory text in this Glossary MAY be reused in IDOCs. However, this text is not intended to authoritatively supersede text of an IDOC in which the Glossary entry is already used. 2.6. Cross-References Some entries contain a parenthetical remark of the form "(See: X.)", where X is a list of other, related terms. Some entries contain a remark of the form "(Compare: X)", where X is a list of terms that either are antonyms of the entry or differ in some other manner worth noting. 2.7. Trademarks All servicemarks and trademarks that appear in this Glossary are used in an editorial fashion and to the benefit of the mark owner, without any intention of infringement. 2.8. The New Punctuation This Glossary uses the "new" or "logical" punctuation style favored by computer programmers, as described by Raymond [Raym]: Programmers use pairs of quotation marks the same way they use pairs of parentheses, i.e., as balanced delimiters. For example, if "Alice sends" is a phrase, and so are "Bill receives" and "Eve listens", then a programmer would write the following sentence: "Alice sends", "Bill receives", and "Eve listens". According to standard American usage, the punctuation in that sentence is incorrect; the continuation commas and the final period should go inside the string quotes, like this: "Alice sends," "Bill receives," and "Eve listens." Shirey Informational [Page 6] RFC 4949 Internet Security Glossary, Version 2 August 2007 However, a programmer would not include a character in a literal string if the character did not belong there, because that could cause an error. For example, suppose a sentence in a draft of a tutorial on the vi editing language looked like this: Then delete one line from the file by typing "dd". A book editor following standard usage might change the sentence to look like this: Then delete one line from the file by typing "dd." However, in the vi language, the dot character repeats the last command accepted. So, if a reader entered "dd.", two lines would be deleted instead of one. Similarly, use of standard American punctuation might cause misunderstanding in entries in this Glossary. Thus, the new punctuation is used here, and we recommend it for IDOCs. 3. Types of Entries Each entry in this Glossary is marked as type I, N, O, or D: 3.1. Type "I": Recommended Definitions of Internet Origin The marking "I" indicates two things: - Origin: "I" (as opposed to "N") means either that the Internet Standards Process or Internet community is authoritative for the definition *or* that the term is sufficiently generic that this Glossary can freely state a definition without contradicting a non-Internet authority (e.g., "attack"). - Recommendation: "I" (as opposed to "O") means that the term and definition are RECOMMENDED for use in IDOCs. However, some "I" entries may be accompanied by a "Usage" note that states a limitation (e.g., "certification"), and IDOCs SHOULD NOT use the defined term outside that limited context. Many "I" entries are proper nouns (e.g., "Internet Protocol") for which the definition is intended only to provide basic information; i.e., the authoritative definition of such terms is found elsewhere. For a proper noun described as an "Internet protocol", please refer to the current edition of "Internet Official Protocol Standards" (Standard 1) for the standardization status of the protocol. Shirey Informational [Page 7] RFC 4949 Internet Security Glossary, Version 2 August 2007 3.2. Type "N": Recommended Definitions of Non-Internet Origin The marking "N" indicates two things: - Origin: "N" (as opposed to "I") means that the entry has a non- Internet basis or origin. - Recommendation: "N" (as opposed to "O") means that the term and definition are RECOMMENDED for use in IDOCs, if they are needed at all in IDOCs. Many of these entries are accompanied by a label that states a context (e.g., "package") or a note that states a limitation (e.g., "data integrity"), and IDOCs SHOULD NOT use the defined term outside that context or limit. Some of the contexts are rarely if ever expected to occur in an IDOC (e.g., "baggage"). In those cases, the listing exists to make Internet authors aware of the non-Internet usage so that they can avoid conflicts with non-Internet documents. 3.3. Type "O": Other Terms and Definitions To Be Noted The marking "O" means that the definition is of non-Internet origin and SHOULD NOT be used in IDOCs *except* in cases where the term is specifically identified as non-Internet. For example, an IDOC might mention "BCA" (see: brand certification authority) or "baggage" as an example of some concept; in that case, the document should specifically say "SET(trademark) BCA" or "SET(trademark) baggage" and include the definition of the term. 3.4. Type "D": Deprecated Terms and Definitions If this Glossary recommends that a term or definition SHOULD NOT be used in IDOCs, then the entry is marked as type "D", and an explanatory note -- "Deprecated Term", "Deprecated Abbreviation", "Deprecated Definition", or "Deprecated Usage" -- is provided. 3.5. Definition Substitutions Some terms have a definition published by a non-Internet authority -- a government (e.g., "object reuse"), an industry (e.g., "Secure Data Exchange"), a national authority (e.g., "Data Encryption Standard"), or an international body (e.g., "data confidentiality") -- that is suitable for use in IDOCs. In those cases, this Glossary marks the definition "N", recommending its use in Internet documents. Other such terms have definitions that are inadequate or inappropriate for IDOCs. For example, a definition might be outdated or too narrow, or it might need clarification by substituting more careful wording (e.g., "authentication exchange") or explanations, using other terms that are defined in this Glossary. In those cases, Shirey Informational [Page 8] RFC 4949 Internet Security Glossary, Version 2 August 2007 this Glossary marks the entry "O", and provides an "I" or "N" entry that precedes, and is intended to supersede, the "O" entry. In some cases where this Glossary provides a definition to supersede an "O" definition, the substitute is intended to subsume the meaning of the "O" entry and not conflict with it. For the term "security service", for example, the "O" definition deals narrowly with only communication services provided by layers in the OSIRM and is inadequate for the full range of IDOC usage, while the new "I" definition provided by this Glossary can be used in more situations and for more kinds of service. However, the "O" definition is also listed so that IDOC authors will be aware of the context in which the term is used more narrowly. When making substitutions, this Glossary attempts to avoid contradicting any non-Internet authority. Still, terminology differs between authorities such as the American Bar Association, OSI, SET, the U.S. DoD, and other authorities; and this Glossary probably is not exactly aligned with any of them. 4. Definitions $ *-property (N) Synonym for "confinement property" in the context of the Bell- LaPadula model. Pronunciation: star property. $ 3DES (N) See: Triple Data Encryption Algorithm. $ A1 computer system (O) /TCSEC/ See: Tutorial under "Trusted Computer System Evaluation Criteria". (Compare: beyond A1.) $ AA (D) See: Deprecated Usage under "attribute authority". $ ABA Guidelines (N) "American Bar Association (ABA) Digital Signature Guidelines" [DSG], a framework of legal principles for using digital signatures and digital certificates in electronic commerce. $ Abstract Syntax Notation One (ASN.1) (N) A standard for describing data objects. [Larm, X680] (See: CMS.) Usage: IDOCs SHOULD use the term "ASN.1" narrowly to describe the notation or language called "Abstract Syntax Notation One". IDOCs MAY use the term more broadly to encompass the notation, its Shirey Informational [Page 9] RFC 4949 Internet Security Glossary, Version 2 August 2007 associated encoding rules (see: BER), and software tools that assist in its use, when the context makes this meaning clear. Tutorial: OSIRM defines computer network functionality in layers. Protocols and data objects at higher layers are abstractly defined to be implemented using protocols and data objects from lower layers. A higher layer may define transfers of abstract objects between computers, and a lower layer may define those transfers concretely as strings of bits. Syntax is needed to specify data formats of abstract objects, and encoding rules are needed to transform abstract objects into bit strings at lower layers. OSI standards use ASN.1 for those specifications and use various encoding rules for those transformations. (See: BER.) In ASN.1, formal names are written without spaces, and separate words in a name are indicated by capitalizing the first letter of each word except the first word. For example, the name of a CRL is "certificateRevocationList". $ ACC (I) See: access control center. $ acceptable risk (I) A risk that is understood and tolerated by a system's user, operator, owner, or accreditor, usually because the cost or difficulty of implementing an effective countermeasure for the associated vulnerability exceeds the expectation of loss. (See: adequate security, risk, "second law" under "Courtney's laws".) $ access 1a. (I) The ability and means to communicate with or otherwise interact with a system to use system resources either to handle information or to gain knowledge of the information the system contains. (Compare: handle.) Usage: The definition is intended to include all types of communication with a system, including one-way communication in either direction. In actual practice, however, passive users might be treated as not having "access" and, therefore, be exempt from most requirements of the system's security policy. (See: "passive user" under "user".) 1b. (O) "Opportunity to make use of an information system (IS) resource." [C4009] 2. (O) /formal model/ "A specific type of interaction between a subject and an object that results in the flow of information from one to the other." [NCS04] Shirey Informational [Page 10] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ Access Certificate for Electronic Services (ACES) (O) A PKI operated by the U.S. Government's General Services Administration in cooperation with industry partners. (See: CAM.) $ access control 1. (I) Protection of system resources against unauthorized access. 2. (I) A process by which use of system resources is regulated according to a security policy and is permitted only by authorized entities (users, programs, processes, or other systems) according to that policy. (See: access, access control service, computer security, discretionary access control, mandatory access control, role-based access control.) 3. (I) /formal model/ Limitations on interactions between subjects and objects in an information system. 4. (O) "The prevention of unauthorized use of a resource, including the prevention of use of a resource in an unauthorized manner." [I7498-2] 5. (O) /U.S. Government/ A system using physical, electronic, or human controls to identify or admit personnel with properly authorized access to a SCIF. $ access control center (ACC) (I) A computer that maintains a database (possibly in the form of an access control matrix) defining the security policy for an access control service, and that acts as a server for clients requesting access control decisions. Tutorial: An ACC is sometimes used in conjunction with a key center to implement access control in a key-distribution system for symmetric cryptography. (See: BLACKER, Kerberos.) $ access control list (ACL) (I) /information system/ A mechanism that implements access control for a system resource by enumerating the system entities that are permitted to access the resource and stating, either implicitly or explicitly, the access modes granted to each entity. (Compare: access control matrix, access list, access profile, capability list.) $ access control matrix (I) A rectangular array of cells, with one row per subject and one column per object. The entry in a cell -- that is, the entry for a particular subject-object pair -- indicates the access mode that the subject is permitted to exercise on the object. Each column is Shirey Informational [Page 11] RFC 4949 Internet Security Glossary, Version 2 August 2007 equivalent to an "access control list" for the object; and each row is equivalent to an "access profile" for the subject. $ access control service (I) A security service that protects against a system entity using a system resource in a way not authorized by the system's security policy. (See: access control, discretionary access control, identity-based security policy, mandatory access control, rule- based security policy.) Tutorial: This service includes protecting against use of a resource in an unauthorized manner by an entity (i.e., a principal) that is authorized to use the resource in some other manner. (See: insider.) The two basic mechanisms for implementing this service are ACLs and tickets. $ access level 1. (D) Synonym for the hierarchical "classification level" in a security level. [C4009] (See: security level.) 2. (D) Synonym for "clearance level". Deprecated Definitions: IDOCs SHOULD NOT use this term with these definitions because they duplicate the meaning of more specific terms. Any IDOC that uses this term SHOULD provide a specific definition for it because access control may be based on many attributes other than classification level and clearance level. $ access list (I) /physical security/ Roster of persons who are authorized to enter a controlled area. (Compare: access control list.) $ access mode (I) A distinct type of data processing operation (e.g., read, write, append, or execute, or a combination of operations) that a subject can potentially perform on an object in an information system. [Huff] (See: read, write.) $ access policy (I) A kind of "security policy". (See: access, access control.) $ access profile (O) Synonym for "capability list". Usage: IDOCs that use this term SHOULD state a definition for it because the definition is not widely known. Shirey Informational [Page 12] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ access right (I) Synonym for "authorization"; emphasizes the possession of the authorization by a system entity. $ accountability (I) The property of a system or system resource that ensures that the actions of a system entity may be traced uniquely to that entity, which can then be held responsible for its actions. [Huff] (See: audit service.) Tutorial: Accountability (a.k.a. individual accountability) typically requires a system ability to positively associate the identity of a user with the time, method, and mode of the user's access to the system. This ability supports detection and subsequent investigation of security breaches. Individual persons who are system users are held accountable for their actions after being notified of the rules of behavior for using the system and the penalties associated with violating those rules. $ accounting See: COMSEC accounting. $ accounting legend code (ALC) (O) /U.S. Government/ Numeric system used to indicate the minimum accounting controls required for items of COMSEC material within the CMCS. [C4009] (See: COMSEC accounting.) $ accreditation (N) An administrative action by which a designated authority declares that an information system is approved to operate in a particular security configuration with a prescribed set of safeguards. [FP102, SP37] (See: certification.) Tutorial: An accreditation is usually based on a technical certification of the system's security mechanisms. To accredit a system, the approving authority must determine that any residual risk is an acceptable risk. Although the terms "certification" and "accreditation" are used more in the U.S. DoD and other U.S. Government agencies than in commercial organizations, the concepts apply any place where managers are required to deal with and accept responsibility for security risks. For example, the American Bar Association is developing accreditation criteria for CAs. $ accreditation boundary (O) Synonym for "security perimeter". [C4009] Shirey Informational [Page 13] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ accreditor (N) A management official who has been designated to have the formal authority to "accredit" an information system, i.e., to authorize the operation of, and the processing of sensitive data in, the system and to accept the residual risk associated with the system. (See: accreditation, residual risk.) $ ACES (O) See: Access Certificate for Electronic Services. $ ACL (I) See: access control list. $ acquirer 1. (O) /SET/ "The financial institution that establishes an account with a merchant and processes payment card authorizations and payments." [SET1] 2. (O) /SET/ "The institution (or its agent) that acquires from the card acceptor the financial data relating to the transaction and initiates that data into an interchange system." [SET2] $ activation data (N) Secret data, other than keys, that is required to access a cryptographic module. (See: CIK. Compare: initialization value.) $ active attack (I) See: secondary definition under "attack". $ active content 1a. (I) Executable software that is bound to a document or other data file and that executes automatically when a user accesses the file, without explicit initiation by the user. (Compare: mobile code.) Tutorial: Active content can be mobile code when its associated file is transferred across a network. 1b. (O) "Electronic documents that can carry out or trigger actions automatically on a computer platform without the intervention of a user. [This technology enables] mobile code associated with a document to execute as the document is rendered." [SP28] $ active user (I) See: secondary definition under "system user". Shirey Informational [Page 14] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ active wiretapping (I) A wiretapping attack that attempts to alter data being communicated or otherwise affect data flow. (See: wiretapping. Compare: active attack, passive wiretapping.) $ add-on security (N) The retrofitting of protection mechanisms, implemented by hardware or software, in an information system after the system has become operational. [FP039] (Compare: baked-in security.) $ adequate security (O) /U.S. DoD/ "Security commensurate with the risk and magnitude of harm resulting from the loss, misuse, or unauthorized access to or modification of information." (See: acceptable risk, residual risk.) $ administrative security 1. (I) Management procedures and constraints to prevent unauthorized access to a system. (See: "third law" under "Courtney's laws", manager, operational security, procedural security, security architecture. Compare: technical security.) Examples: Clear delineation and separation of duties; configuration control. Usage: Administrative security is usually understood to consist of methods and mechanisms that are implemented and executed primarily by people, rather than by automated systems. 2. (O) "The management constraints, operational procedures, accountability procedures, and supplemental controls established to provide an acceptable level of protection for sensitive data." [FP039] $ administrator 1. (O) /Common Criteria/ A person that is responsible for configuring, maintaining, and administering the TOE in a correct manner for maximum security. (See: administrative security.) 2. (O) /ITSEC/ A person in contact with the TOE, who is responsible for maintaining its operational capability. $ Advanced Encryption Standard (AES) (N) A U.S. Government standard [FP197] (the successor to DES) that (a) specifies "the AES algorithm", which is a symmetric block cipher that is based on Rijndael and uses key sizes of 128, 192, or 256 bits to operate on a 128-bit block, and (b) states policy for using that algorithm to protect unclassified, sensitive data. Shirey Informational [Page 15] RFC 4949 Internet Security Glossary, Version 2 August 2007 Tutorial: Rijndael was designed to handle additional block sizes and key lengths that were not adopted in the AES. Rijndael was selected by NIST through a public competition that was held to find a successor to the DEA; the other finalists were MARS, RC6, Serpent, and Twofish. $ adversary 1. (I) An entity that attacks a system. (Compare: cracker, intruder, hacker.) 2. (I) An entity that is a threat to a system. $ AES (N) See: Advanced Encryption Standard. $ Affirm (O) A formal methodology, language, and integrated set of software tools developed at the University of Southern California's Information Sciences Institute for specifying, coding, and verifying software to produce correct and reliable programs. [Cheh] $ aggregation (I) A circumstance in which a collection of information items is required to be classified at a higher security level than any of the items is classified individually. (See: classification.) $ AH (I) See: Authentication Header $ air gap (I) An interface between two systems at which (a) they are not connected physically and (b) any logical connection is not automated (i.e., data is transferred through the interface only manually, under human control). (See: sneaker net. Compare: gateway.) Example: Computer A and computer B are on opposite sides of a room. To move data from A to B, a person carries a disk across the room. If A and B operate in different security domains, then moving data across the air gap may involve an upgrade or downgrade operation. $ ALC (O) See: accounting legend code. Shirey Informational [Page 16] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ algorithm (I) A finite set of step-by-step instructions for a problem- solving or computation procedure, especially one that can be implemented by a computer. (See: cryptographic algorithm.) $ alias (I) A name that an entity uses in place of its real name, usually for the purpose of either anonymity or masquerade. $ Alice and Bob (I) The parties that are most often called upon to illustrate the operation of bipartite security protocols. These and other dramatis personae are listed by Schneier [Schn]. $ American National Standards Institute (ANSI) (N) A private, not-for-profit association that administers U.S. private-sector voluntary standards. Tutorial: ANSI has approximately 1,000 member organizations, including equipment users, manufacturers, and others. These include commercial firms, governmental agencies, and other institutions and international entities. ANSI is the sole U.S. representative to (a) ISO and (b) (via the U.S. National Committee) the International Electrotechnical Commission (IEC), which are the two major, non-treaty, international standards organizations. ANSI provides a forum for ANSI-accredited standards development groups. Among those groups, the following are especially relevant to Internet security: - International Committee for Information Technology Standardization (INCITS) (formerly X3): Primary U.S. focus of standardization in information and communications technologies, encompassing storage, processing, transfer, display, management, organization, and retrieval of information. Example: [A3092]. - Accredited Standards Committee X9: Develops, establishes, maintains, and promotes standards for the financial services industry. Example: [A9009]. - Alliance for Telecommunications Industry Solutions (ATIS): Develops standards, specifications, guidelines, requirements, technical reports, industry processes, and verification tests for interoperability and reliability of telecommunications networks, equipment, and software. Example: [A1523]. Shirey Informational [Page 17] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ American Standard Code for Information Interchange (ASCII) (N) A scheme that encodes 128 specified characters -- the numbers 0-9, the letters a-z and A-Z, some basic punctuation symbols, some control codes that originated with Teletype machines, and a blank space -- into the 7-bit binary integers. Forms the basis of the character set representations used in most computers and many Internet standards. [FP001] (See: code.) $ Anderson report (O) A 1972 study of computer security that was written by James P. Anderson for the U.S. Air Force [Ande]. Tutorial: Anderson collaborated with a panel of experts to study Air Force requirements for multilevel security. The study recommended research and development that was urgently needed to provide secure information processing for command and control systems and support systems. The report introduced the reference monitor concept and provided development impetus for computer and network security technology. However, many of the security problems that the 1972 report called "current" still plague information systems today. $ anomaly detection (I) An intrusion detection method that searches for activity that is different from the normal behavior of system entities and system resources. (See: IDS. Compare: misuse detection.) $ anonymity (I) The condition of an identity being unknown or concealed. (See: alias, anonymizer, anonymous credential, anonymous login, identity, onion routing, persona certificate. Compare: privacy.) Tutorial: An application may require security services that maintain anonymity of users or other system entities, perhaps to preserve their privacy or hide them from attack. To hide an entity's real name, an alias may be used; for example, a financial institution may assign account numbers. Parties to transactions can thus remain relatively anonymous, but can also accept the transactions as legitimate. Real names of the parties cannot be easily determined by observers of the transactions, but an authorized third party may be able to map an alias to a real name, such as by presenting the institution with a court order. In other applications, anonymous entities may be completely untraceable. $ anonymizer (I) An internetwork service, usually provided via a proxy server, that provides anonymity and privacy for clients. That is, the service enables a client to access servers (a) without allowing Shirey Informational [Page 18] RFC 4949 Internet Security Glossary, Version 2 August 2007 anyone to gather information about which servers the client accesses and (b) without allowing the accessed servers to gather information about the client, such as its IP address. $ anonymous credential (D) /U.S. Government/ A credential that (a) can be used to authenticate a person as having a specific attribute or being a member of a specific group (e.g., military veterans or U.S. citizens) but (b) does not reveal the individual identity of the person that presents the credential. [M0404] (See: anonymity.) Deprecated Term: IDOCs SHOULD NOT use this term; it mixes concepts in a potentially misleading way. For example, when the credential is an X.509 certificate, the term could be misunderstood to mean that the certificate was signed by a CA that has a persona certificate. Instead, use "attribute certificate", "organizational certificate", or "persona certificate" depending on what is meant, and provide additional explanations as needed. $ anonymous login (I) An access control feature (actually, an access control vulnerability) in many Internet hosts that enables users to gain access to general-purpose or public services and resources of a host (such as allowing any user to transfer data using FTP) without having a pre-established, identity-specific account (i.e., user name and password). (See: anonymity.) Tutorial: This feature exposes a system to more threats than when all the users are known, pre-registered entities that are individually accountable for their actions. A user logs in using a special, publicly known user name (e.g., "anonymous", "guest", or "ftp"). To use the public login name, the user is not required to know a secret password and may not be required to input anything at all except the name. In other cases, to complete the normal sequence of steps in a login protocol, the system may require the user to input a matching, publicly known password (such as "anonymous") or may ask the user for an e-mail address or some other arbitrary character string. $ ANSI (N) See: American National Standards Institute. $ anti-jam (N) "Measures ensuring that transmitted information can be received despite deliberate jamming attempts." [C4009] (See: electronic security, frequency hopping, jam, spread spectrum.) Shirey Informational [Page 19] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ apex trust anchor (N) The trust anchor that is superior to all other trust anchors in a particular system or context. (See: trust anchor, top CA.) $ API (I) See: application programming interface. $ APOP (I) See: POP3 APOP. $ Application Layer See: Internet Protocol Suite, OSIRM. $ application program (I) A computer program that performs a specific function directly for a user (as opposed to a program that is part of a computer operating system and exists to perform functions in support of application programs). $ architecture (I) See: security architecture, system architecture. $ archive 1a. (I) /noun/ A collection of data that is stored for a relatively long period of time for historical and other purposes, such as to support audit service, availability service, or system integrity service. (Compare: backup, repository.) 1b. (I) /verb/ To store data in such a way as to create an archive. (Compare: back up.) Tutorial: A digital signature may need to be verified many years after the signing occurs. The CA -- the one that issued the certificate containing the public key needed to verify that signature -- may not stay in operation that long. So every CA needs to provide for long-term storage of the information needed to verify the signatures of those to whom it issues certificates. $ ARPANET (I) Advanced Research Projects Agency (ARPA) Network, a pioneer packet-switched network that (a) was designed, implemented, operated, and maintained by BBN from January 1969 until July 1975 under contract to the U.S. Government; (b) led to the development of today's Internet; and (c) was decommissioned in June 1990. [B4799, Hafn] $ ASCII (N) See: American Standard Code for Information Interchange. Shirey Informational [Page 20] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ ASN.1 (N) See: Abstract Syntax Notation One. $ asset (I) A system resource that is (a) required to be protected by an information system's security policy, (b) intended to be protected by a countermeasure, or (c) required for a system's mission. $ association (I) A cooperative relationship between system entities, usually for the purpose of transferring information between them. (See: security association.) $ assurance See: security assurance. $ assurance level (N) A rank on a hierarchical scale that judges the confidence someone can have that a TOE adequately fulfills stated security requirements. (See: assurance, certificate policy, EAL, TCSEC.) Example: U.S. Government guidance [M0404] describes four assurance levels for identity authentication, where each level "describes the [U.S. Federal Government] agency's degree of certainty that the user has presented [a credential] that refers to [the user's] identity." In that guidance, assurance is defined as (a) "the degree of confidence in the vetting process used to establish the identity of the individual to whom the credential was issued" and (b) "the degree of confidence that the individual who uses the credential is the individual to whom the credential was issued." The four levels are described as follows: - Level 1: Little or no confidence in the asserted identity. - Level 2: Some confidence in the asserted identity. - Level 3: High confidence in the asserted identity. - Level 4: Very high confidence in the asserted identity. Standards for determining these levels are provided in a NIST publication [SP12]. However, as noted there, an assurance level is "a degree of confidence, not a true measure of how secure the system actually is. This distinction is necessary because it is extremely difficult -- and in many cases, virtually impossible -- to know exactly how secure a system is." $ asymmetric cryptography (I) A modern branch of cryptography (popularly known as "public- key cryptography") in which the algorithms use a pair of keys (a public key and a private key) and use a different component of the pair for each of two counterpart cryptographic operations (e.g., Shirey Informational [Page 21] RFC 4949 Internet Security Glossary, Version 2 August 2007 encryption and decryption, or signature creation and signature verification). (See: key pair, symmetric cryptography.) Tutorial: Asymmetric algorithms have key management advantages over equivalently strong symmetric ones. First, one key of the pair need not be known by anyone but its owner; so it can more easily be kept secret. Second, although the other key is shared by all entities that use the algorithm, that key need not be kept secret from other, non-using entities; thus, the key-distribution part of key management can be done more easily. Asymmetric cryptography can be used to create algorithms for encryption, digital signature, and key agreement: - In an asymmetric encryption algorithm (e.g., "RSA"), when Alice wants to ensure confidentiality for data she sends to Bob, she encrypts the data with a public key provided by Bob. Only Bob has the matching private key that is needed to decrypt the data. (Compare: seal.) - In an asymmetric digital signature algorithm (e.g., "DSA"), when Alice wants to ensure data integrity or provide authentication for data she sends to Bob, she uses her private key to sign the data (i.e., create a digital signature based on the data). To verify the signature, Bob uses the matching public key that Alice has provided. - In an asymmetric key-agreement algorithm (e.g., "Diffie- Hellman-Merkle"), Alice and Bob each send their own public key to the other party. Then each uses their own private key and the other's public key to compute the new key value. $ asymmetric key (I) A cryptographic key that is used in an asymmetric cryptographic algorithm. (See: asymmetric cryptography, private key, public key.) $ ATIS (N) See: "Alliance for Telecommunications Industry Solutions" under "ANSI". $ attack 1. (I) An intentional act by which an entity attempts to evade security services and violate the security policy of a system. That is, an actual assault on system security that derives from an intelligent threat. (See: penetration, violation, vulnerability.) 2. (I) A method or technique used in an assault (e.g., masquerade). (See: blind attack, distributed attack.) Shirey Informational [Page 22] RFC 4949 Internet Security Glossary, Version 2 August 2007 Tutorial: Attacks can be characterized according to intent: - An "active attack" attempts to alter system resources or affect their operation. - A "passive attack" attempts to learn or make use of information from a system but does not affect system resources of that system. (See: wiretapping.) The object of a passive attack might be to obtain data that is needed for an off-line attack. - An "off-line attack" is one in which the attacker obtains data from the target system and then analyzes the data on a different system of the attacker's own choosing, possibly in preparation for a second stage of attack on the target. Attacks can be characterized according to point of initiation: - An "inside attack" is one that is initiated by an entity inside the security perimeter (an "insider"), i.e., an entity that is authorized to access system resources but uses them in a way not approved by the party that granted the authorization. - An "outside attack" is initiated from outside the security perimeter, by an unauthorized or illegitimate user of the system (an "outsider"). In the Internet, potential outside attackers range from amateur pranksters to organized criminals, international terrorists, and hostile governments. Attacks can be characterized according to method of delivery: - In a "direct attack", the attacker addresses attacking packets to the intended victim(s). - In an "indirect attack", the attacker addresses packets to a third party, and the packets either have the address(es) of the intended victim(s) as their source address(es) or indicate the intended victim(s) in some other way. The third party responds by sending one or more attacking packets to the intended victims. The attacker can use third parties as attack amplifiers by providing a broadcast address as the victim address (e.g., "smurf attack"). (See: reflector attack. Compare: reflection attack, replay attack.) Shirey Informational [Page 23] RFC 4949 Internet Security Glossary, Version 2 August 2007 The term "attack" relates to some other basic security terms as shown in the following diagram: + - - - - - - - - - - - - + + - - - - + + - - - - - - - - - - -+ | An Attack: | |Counter- | | A System Resource: | | i.e., A Threat Action | | measure | | Target of the Attack | | +----------+ | | | | +-----------------+ | | | Attacker |<==================||<========= | | | | i.e., | Passive | | | | | Vulnerability | | | | A Threat |<=================>||<========> | | | | Agent | or Active | | | | +-------|||-------+ | | +----------+ Attack | | | | VVV | | | | | | Threat Consequences | + - - - - - - - - - - - - + + - - - - + + - - - - - - - - - - -+ $ attack potential (I) The perceived likelihood of success should an attack be launched, expressed in terms of the attacker's ability (i.e., expertise and resources) and motivation. (Compare: threat, risk.) $ attack sensing, warning, and response (I) A set of security services that cooperate with audit service to detect and react to indications of threat actions, including both inside and outside attacks. (See: indicator.) $ attack tree (I) A branching, hierarchical data structure that represents a set of potential approaches to achieving an event in which system security is penetrated or compromised in a specified way. [Moor] Tutorial: Attack trees are special cases of fault trees. The security incident that is the goal of the attack is represented as the root node of the tree, and the ways that an attacker could reach that goal are iteratively and incrementally represented as branches and subnodes of the tree. Each subnode defines a subgoal, and each subgoal may have its own set of further subgoals, etc. The final nodes on the paths outward from the root, i.e., the leaf nodes, represent different ways to initiate an attack. Each node other than a leaf is either an AND-node or an OR-node. To achieve the goal represented by an AND-node, the subgoals represented by all of that node's subnodes must be achieved; and for an OR-node, at least one of the subgoals must be achieved. Branches can be labeled with values representing difficulty, cost, or other attack attributes, so that alternative attacks can be compared. Shirey Informational [Page 24] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ attribute (N) Information of a particular type concerning an identifiable system entity or object. An "attribute type" is the component of an attribute that indicates the class of information given by the attribute; and an "attribute value" is a particular instance of the class of information indicated by an attribute type. (See: attribute certificate.) $ attribute authority (AA) 1. (N) A CA that issues attribute certificates. 2. (O) "An authority [that] assigns privileges by issuing attribute certificates." [X509] Deprecated Usage: The abbreviation "AA" SHOULD NOT be used in an IDOC unless it is first defined in the IDOC. $ attribute certificate 1. (I) A digital certificate that binds a set of descriptive data items, other than a public key, either directly to a subject name or to the identifier of another certificate that is a public-key certificate. (See: capability token.) 2. (O) "A data structure, digitally signed by an [a]ttribute [a]uthority, that binds some attribute values with identification information about its holder." [X509] Tutorial: A public-key certificate binds a subject name to a public key value, along with information needed to perform certain cryptographic functions using that key. Other attributes of a subject, such as a security clearance, may be certified in a separate kind of digital certificate, called an attribute certificate. A subject may have multiple attribute certificates associated with its name or with each of its public-key certificates. An attribute certificate might be issued to a subject in the following situations: - Different lifetimes: When the lifetime of an attribute binding is shorter than that of the related public-key certificate, or when it is desirable not to need to revoke a subject's public key just to revoke an attribute. - Different authorities: When the authority responsible for the attributes is different than the one that issues the public-key certificate for the subject. (There is no requirement that an attribute certificate be issued by the same CA that issued the associated public-key certificate.) Shirey Informational [Page 25] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ audit See: security audit. $ audit log (I) Synonym for "security audit trail". $ audit service (I) A security service that records information needed to establish accountability for system events and for the actions of system entities that cause them. (See: security audit.) $ audit trail (I) See: security audit trail. $ AUTH (I) See: POP3 AUTH. $ authenticate (I) Verify (i.e., establish the truth of) an attribute value claimed by or for a system entity or system resource. (See: authentication, validate vs. verify, "relationship between data integrity service and authentication services" under "data integrity service".) Deprecated Usage: In general English usage, this term is used with the meaning "to prove genuine" (e.g., an art expert authenticates a Michelangelo painting); but IDOCs should restrict usage as follows: - IDOCs SHOULD NOT use this term to refer to proving or checking that data has not been changed, destroyed, or lost in an unauthorized or accidental manner. Instead, use "verify". - IDOCs SHOULD NOT use this term to refer to proving the truth or accuracy of a fact or value such as a digital signature. Instead, use "verify". - IDOCs SHOULD NOT use this term to refer to establishing the soundness or correctness of a construct, such as a digital certificate. Instead, use "validate". $ authentication (I) The process of verifying a claim that a system entity or system resource has a certain attribute value. (See: attribute, authenticate, authentication exchange, authentication information, credential, data origin authentication, peer entity authentication, "relationship between data integrity service and authentication services" under "data integrity service", simple authentication, strong authentication, verification, X.509.) Shirey Informational [Page 26] RFC 4949 Internet Security Glossary, Version 2 August 2007 Tutorial: Security services frequently depend on authentication of the identity of users, but authentication may involve any type of attribute that is recognized by a system. A claim may be made by a subject about itself (e.g., at login, a user typically asserts its identity) or a claim may be made on behalf of a subject or object by some other system entity (e.g., a user may claim that a data object originates from a specific source, or that a data object is classified at a specific security level). An authentication process consists of two basic steps: - Identification step: Presenting the claimed attribute value (e.g., a user identifier) to the authentication subsystem. - Verification step: Presenting or generating authentication information (e.g., a value signed with a private key) that acts as evidence to prove the binding between the attribute and that for which it is claimed. (See: verification.) $ authentication code (D) Synonym for a checksum based on cryptography. (Compare: Data Authentication Code, Message Authentication Code.) Deprecated Term: IDOCs SHOULD NOT use this uncapitalized term as a synonym for any kind of checksum, regardless of whether or not the checksum is cryptographic. Instead, use "checksum", "Data Authentication Code", "error detection code", "hash", "keyed hash", "Message Authentication Code", "protected checksum", or some other recommended term, depending on what is meant. The term mixes concepts in a potentially misleading way. The word "authentication" is misleading because the checksum may be used to perform a data integrity function rather than a data origin authentication function. $ authentication exchange 1. (I) A mechanism to verify the identity of an entity by means of information exchange. 2. (O) "A mechanism intended to ensure the identity of an entity by means of information exchange." [I7498-2] $ Authentication Header (AH) (I) An Internet protocol [R2402, R4302] designed to provide connectionless data integrity service and connectionless data origin authentication service for IP datagrams, and (optionally) to provide partial sequence integrity and protection against replay attacks. (See: IPsec. Compare: ESP.) Shirey Informational [Page 27] RFC 4949 Internet Security Glossary, Version 2 August 2007 Tutorial: Replay protection may be selected by the receiver when a security association is established. AH authenticates the upper- layer PDU that is carried as an IP SDU, and also authenticates as much of the IP PCI (i.e., the IP header) as possible. However, some IP header fields may change in transit, and the value of these fields, when the packet arrives at the receiver, may not be predictable by the sender. Thus, the values of such fields cannot be protected end-to-end by AH; protection of the IP header by AH is only partial when such fields are present. AH may be used alone, or in combination with the ESP, or in a nested fashion with tunneling. Security services can be provided between a pair of communicating hosts, between a pair of communicating security gateways, or between a host and a gateway. ESP can provide nearly the same security services as AH, and ESP can also provide data confidentiality service. The main difference between authentication services provided by ESP and AH is the extent of the coverage; ESP does not protect IP header fields unless they are encapsulated by AH. $ authentication information (I) Information used to verify an identity claimed by or for an entity. (See: authentication, credential, user. Compare: identification information.) Tutorial: Authentication information may exist as, or be derived from, one of the following: (a) Something the entity knows (see: password); (b) something the entity possesses (see: token); (c) something the entity is (see: biometric authentication). $ authentication service (I) A security service that verifies an identity claimed by or for an entity. (See: authentication.) Tutorial: In a network, there are two general forms of authentication service: data origin authentication service and peer entity authentication service. $ authenticity (I) The property of being genuine and able to be verified and be trusted. (See: authenticate, authentication, validate vs. verify.) $ authority (D) /PKI/ "An entity [that is] responsible for the issuance of certificates." [X509] Shirey Informational [Page 28] RFC 4949 Internet Security Glossary, Version 2 August 2007 Deprecated Usage: IDOCs SHOULD NOT use this term as a synonym for attribute authority, certification authority, registration authority, or similar terms; the shortened form may cause confusion. Instead, use the full term at the first instance of usage and then, if it is necessary to shorten text, use AA, CA, RA, and other abbreviations defined in this Glossary. $ authority certificate (D) "A certificate issued to an authority (e.g. either to a certification authority or to an attribute authority)." [X509] (See: authority.) Deprecated Term: IDOCs SHOULD NOT use this term because it is ambiguous. Instead, use the full term "certification authority certificate", "attribute authority certificate", "registration authority certificate", etc. at the first instance of usage and then, if it is necessary to shorten text, use AA, CA, RA, and other abbreviations defined in this Glossary. $ Authority Information Access extension (I) The private extension defined by PKIX for X.509 certificates to indicate "how to access CA information and services for the issuer of the certificate in which the extension appears. Information and services may include on-line validation services and CA policy data." [R3280] (See: private extension.) $ authorization 1a. (I) An approval that is granted to a system entity to access a system resource. (Compare: permission, privilege.) Usage: Some synonyms are "permission" and "privilege". Specific terms are preferred in certain contexts: - /PKI/ "Authorization" SHOULD be used, to align with "certification authority" in the standard [X509]. - /role-based access control/ "Permission" SHOULD be used, to align with the standard [ANSI]. - /computer operating systems/ "Privilege" SHOULD be used, to align with the literature. (See: privileged process, privileged user.) Tutorial: The semantics and granularity of authorizations depend on the application and implementation (see: "first law" under "Courtney's laws"). An authorization may specify a particular access mode -- such as read, write, or execute -- for one or more system resources. 1b. (I) A process for granting approval to a system entity to access a system resource. Shirey Informational [Page 29] RFC 4949 Internet Security Glossary, Version 2 August 2007 2. (O) /SET/ "The process by which a properly appointed person or persons grants permission to perform some action on behalf of an organization. This process assesses transaction risk, confirms that a given transaction does not raise the account holder's debt above the account's credit limit, and reserves the specified amount of credit. (When a merchant obtains authorization, payment for the authorized amount is guaranteed -- provided, of course, that the merchant followed the rules associated with the authorization process.)" [SET2] $ authorization credential (I) See: /access control/ under "credential". $ authorize (I) Grant an authorization to a system entity. $ authorized user (I) /access control/ A system entity that accesses a system resource for which the entity has received an authorization. (Compare: insider, outsider, unauthorized user.) Deprecated Usage: IDOCs that use this term SHOULD state a definition for it because the term is used in many ways and could easily be misunderstood. $ automated information system See: information system. $ availability 1. (I) The property of a system or a system resource being accessible, or usable or operational upon demand, by an authorized system entity, according to performance specifications for the system; i.e., a system is available if it provides services according to the system design whenever users request them. (See: critical, denial of service. Compare: precedence, reliability, survivability.) 2. (O) "The property of being accessible and usable upon demand by an authorized entity." [I7498-2] 3. (D) "Timely, reliable access to data and information services for authorized users." [C4009] Deprecated Definition: IDOCs SHOULD NOT use the term with definition 3; the definition mixes "availability" with "reliability", which is a different property. (See: reliability.) Shirey Informational [Page 30] RFC 4949 Internet Security Glossary, Version 2 August 2007 Tutorial: Availability requirements can be specified by quantitative metrics, but sometimes are stated qualitatively, such as in the following: - "Flexible tolerance for delay" may mean that brief system outages do not endanger mission accomplishment, but extended outages may endanger the mission. - "Minimum tolerance for delay" may mean that mission accomplishment requires the system to provide requested services in a short time. $ availability service (I) A security service that protects a system to ensure its availability. Tutorial: This service addresses the security concerns raised by denial-of-service attacks. It depends on proper management and control of system resources, and thus depends on access control service and other security services. $ avoidance (I) See: secondary definition under "security". $ B1, B2, or B3 computer system (O) /TCSEC/ See: Tutorial under "Trusted Computer System Evaluation Criteria". $ back door 1. (I) /COMPUSEC/ A computer system feature -- which may be (a) an unintentional flaw, (b) a mechanism deliberately installed by the system's creator, or (c) a mechanism surreptitiously installed by an intruder -- that provides access to a system resource by other than the usual procedure and usually is hidden or otherwise not well-known. (See: maintenance hook. Compare: Trojan Horse.) Example: A way to access a computer other than through a normal login. Such an access path is not necessarily designed with malicious intent; operating systems sometimes are shipped by the manufacturer with hidden accounts intended for use by field service technicians or the vendor's maintenance programmers. 2. (I) /cryptography/ A feature of a cryptographic system that makes it easily possible to break or circumvent the protection that the system is designed to provide. Example: A feature that makes it possible to decrypt cipher text much more quickly than by brute-force cryptanalysis, without having prior knowledge of the decryption key. Shirey Informational [Page 31] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ back up (I) /verb/ Create a reserve copy of data or, more generally, provide alternate means to perform system functions despite loss of system resources. (See: contingency plan. Compare: archive.) $ backup (I) /noun or adjective/ Refers to alternate means of performing system functions despite loss of system resources. (See: contingency plan). Example: A reserve copy of data, preferably one that is stored separately from the original, for use if the original becomes lost or damaged. (Compare: archive.) $ bagbiter (D) /slang/ "An entity, such as a program or a computer, that fails to work or that works in a remarkably clumsy manner. A person who has caused some trouble, inadvertently or otherwise, typically by failing to program the computer properly." [NCSSG] (See: flaw.) Deprecated Term: It is likely that other cultures use different metaphors for these concepts. Therefore, to avoid international misunderstanding, IDOCs SHOULD NOT use this term. (See: Deprecated Usage under "Green Book".) $ baggage (O) /SET/ An "opaque encrypted tuple, which is included in a SET message but appended as external data to the PKCS encapsulated data. This avoids superencryption of the previously encrypted tuple, but guarantees linkage with the PKCS portion of the message." [SET2] Deprecated Usage: IDOCs SHOULD NOT use this term to describe a data element, except in the form "SET(trademark) baggage" with the meaning given above. $ baked-in security (D) The inclusion of security mechanisms in an information system beginning at an early point in the system's lifecycle, i.e., during the design phase, or at least early in the implementation phase. (Compare: add-on security.) Deprecated Term: It is likely that other cultures use different metaphors for this concept. Therefore, to avoid international misunderstanding, IDOCs SHOULD NOT use this term (unless they also provide a definition like this one). (See: Deprecated Usage under "Green Book".) Shirey Informational [Page 32] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ bandwidth (I) The total width of the frequency band that is available to or used by a communication channel; usually expressed in Hertz (Hz). (RFC 3753) (Compare: channel capacity.) $ bank identification number (BIN) 1. (O) The digits of a credit card number that identify the issuing bank. (See: primary account number.) 2. (O) /SET/ The first six digits of a primary account number. $ Basic Encoding Rules (BER) (I) A standard for representing ASN.1 data types as strings of octets. [X690] (See: Distinguished Encoding Rules.) Deprecated Usage: Sometimes incorrectly treated as part of ASN.1. However, ASN.1 properly refers only to a syntax description language, and not to the encoding rules for the language. $ Basic Security Option (I) See: secondary definition under "IPSO". $ bastion host (I) A strongly protected computer that is in a network protected by a firewall (or is part of a firewall) and is the only host (or one of only a few) in the network that can be directly accessed from networks on the other side of the firewall. (See: firewall.) Tutorial: Filtering routers in a firewall typically restrict traffic from the outside network to reaching just one host, the bastion host, which usually is part of the firewall. Since only this one host can be directly attacked, only this one host needs to be very strongly protected, so security can be maintained more easily and less expensively. However, to allow legitimate internal and external users to access application resources through the firewall, higher-layer protocols and services need to be relayed and forwarded by the bastion host. Some services (e.g., DNS and SMTP) have forwarding built in; other services (e.g., TELNET and FTP) require a proxy server on the bastion host. $ BBN Technologies Corp. (BBN) (O) The research-and-development company (originally called Bolt Baranek and Newman, Inc.) that built the ARPANET. $ BCA (O) See: brand certification authority. Shirey Informational [Page 33] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ BCR (O) See: BLACK/Crypto/RED. $ BCI (O) See: brand CRL identifier. $ Bell-LaPadula model (N) A formal, mathematical, state-transition model of confidentiality policy for multilevel-secure computer systems [Bell]. (Compare: Biba model, Brewer-Nash model.) Tutorial: The model, devised by David Bell and Leonard LaPadula at The MITRE Corporation in 1973, characterizes computer system elements as subjects and objects. To determine whether or not a subject is authorized for a particular access mode on an object, the clearance of the subject is compared to the classification of the object. The model defines the notion of a "secure state", in which the only permitted access modes of subjects to objects are in accordance with a specified security policy. It is proven that each state transition preserves security by moving from secure state to secure state, thereby proving that the system is secure. In this model, a multilevel-secure system satisfies several rules, including the "confinement property" (a.k.a. the "*-property"), the "simple security property", and the "tranquility property". $ benign 1. (N) /COMSEC/ "Condition of cryptographic data [such] that [the data] cannot be compromised by human access [to the data]." [C4009] 2. (O) /COMPUSEC/ See: secondary definition under "trust". $ benign fill (N) Process by which keying material is generated, distributed, and placed into an ECU without exposure to any human or other system entity, except the cryptographic module that consumes and uses the material. (See: benign.) $ BER (I) See: Basic Encoding Rules. $ beyond A1 1. (O) /formal/ A level of security assurance that is beyond the highest level (level A1) of criteria specified by the TCSEC. (See: Tutorial under "Trusted Computer System Evaluation Criteria".) Shirey Informational [Page 34] RFC 4949 Internet Security Glossary, Version 2 August 2007 2. (O) /informal/ A level of trust so high that it is beyond state-of-the-art technology; i.e., it cannot be provided or verified by currently available assurance methods, and especially not by currently available formal methods. $ Biba integrity (N) Synonym for "source integrity". $ Biba model (N) A formal, mathematical, state-transition model of integrity policy for multilevel-secure computer systems [Biba]. (See: source integrity. Compare: Bell-LaPadula model.) Tutorial: This model for integrity control is analogous to the Bell-LaPadula model for confidentiality control. Each subject and object is assigned an integrity level and, to determine whether or not a subject is authorized for a particular access mode on an object, the integrity level of the subject is compared to that of the object. The model prohibits the changing of information in an object by a subject with a lesser or incomparable level. The rules of the Biba model are duals of the corresponding rules in the Bell-LaPadula model. $ billet (N) "A personnel position or assignment that may be filled by one person." [JCP1] (Compare: principal, role, user.) Tutorial: In an organization, a "billet" is a populational position, of which there is exactly one instance; but a "role" is functional position, of which there can be multiple instances. System entities are in one-to-one relationships with their billets, but may be in many-to-one and one-to-many relationships with their roles. $ BIN (O) See: bank identification number. $ bind (I) To inseparably associate by applying some security mechanism. Example: A CA creates a public-key certificate by using a digital signature to bind together (a) a subject name, (b) a public key, and usually (c) some additional data items (e.g., "X.509 public- key certificate"). $ biometric authentication (I) A method of generating authentication information for a person by digitizing measurements of a physical or behavioral Shirey Informational [Page 35] RFC 4949 Internet Security Glossary, Version 2 August 2007 characteristic, such as a fingerprint, hand shape, retina pattern, voiceprint, handwriting style, or face. $ birthday attack (I) A class of attacks against cryptographic functions, including both encryption functions and hash functions. The attacks take advantage of a statistical property: Given a cryptographic function having an N-bit output, the probability is greater than 1/2 that for 2**(N/2) randomly chosen inputs, the function will produce at least two outputs that are identical. (See: Tutorial under "hash function".) Derivation: From the somewhat surprising fact (often called the "birthday paradox") that although there are 365 days in a year, the probability is greater than 1/2 that two of more people share the same birthday in any randomly chosen group of 23 people. Birthday attacks enable an adversary to find two inputs for which a cryptographic function produces the same cipher text (or find two inputs for which a hash functions produces the same hash result) much faster than a brute-force attack can; and a clever adversary can use such a capability to create considerable mischief. However, no birthday attack can enable an adversary to decrypt a given cipher text (or find a hash input that results in a given hash result) any faster than a brute-force attack can. $ bit (I) A contraction of the term "binary digit"; the smallest unit of information storage, which has two possible states or values. The values usually are represented by the symbols "0" (zero) and "1" (one). (See: block, byte, nibble, word.) $ bit string (I) A sequence of bits, each of which is either "0" or "1". $ BLACK 1. (N) Designation for data that consists only of cipher text, and for information system equipment items or facilities that handle only cipher text. Example: "BLACK key". (See: BCR, color change, RED/BLACK separation. Compare: RED.) 2. (O) /U.S. Government/ "Designation applied to information systems, and to associated areas, circuits, components, and equipment, in which national security information is encrypted or is not processed." [C4009] 3. (D) Any data that can be disclosed without harm. Shirey Informational [Page 36] RFC 4949 Internet Security Glossary, Version 2 August 2007 Deprecated Definition: IDOCs SHOULD NOT use the term with definition 3 because the definition is ambiguous with regard to whether or not the data is protected. $ BLACK/Crypto/RED (BCR) (N) An experimental, end-to-end, network packet encryption system developed in a working prototype form by BBN and the Collins Radio division of Rockwell Corporation in the 1975-1980 time frame for the U.S. DoD. BCR was the first network security system to support TCP/IP traffic, and it incorporated the first DES chips that were validated by the U.S. National Bureau of Standards (now called NIST). BCR also was the first to use a KDC and an ACC to manage connections. $ BLACK key (N) A key that is protected with a key-encrypting key and that must be decrypted before use. (See: BLACK. Compare: RED key.) $ BLACKER (O) An end-to-end encryption system for computer data networks that was developed by the U.S. DoD in the 1980s to provide host- to-host data confidentiality service for datagrams at OSIRM Layer 3. [Weis] (Compare: CANEWARE, IPsec.) Tutorial: Each user host connects to its own bump-in-the-wire encryption device called a BLACKER Front End (BFE, TSEC/KI-111), through which the host connects to the subnetwork. The system also includes two types of centralized devices: one or more KDCs connect to the subnetwork and communicate with assigned sets of BFEs, and one or more ACCs connect to the subnetwork and communicate with assigned KDCs. BLACKER uses only symmetric encryption. A KDC distributes session keys to BFE pairs as authorized by an ACC. Each ACC maintains a database for a set of BFEs, and the database determines which pairs from that set (i.e., which pairs of user hosts behind the BFEs) are authorized to communicate and at what security levels. The BLACKER system is MLS in three ways: (a) The BFEs form a security perimeter around a subnetwork, separating user hosts from the subnetwork, so that the subnetwork can operate at a different security level (possibly a lower, less expensive level) than the hosts. (b) The BLACKER components are trusted to separate datagrams of different security levels, so that each datagram of a given security level can be received only by a host that is authorized for that security level; and thus BLACKER can separate host communities that operate at different security levels. (c) The host side of a BFE is itself MLS and can recognize a security label on each packet, so that an MLS user host can be authorized Shirey Informational [Page 37] RFC 4949 Internet Security Glossary, Version 2 August 2007 to successively transmit datagrams that are labeled with different security levels. $ blind attack (I) A type of network-based attack method that does not require the attacking entity to receive data traffic from the attacked entity; i.e., the attacker does not need to "see" data packets sent by the victim. Example: SYN flood. Tutorial: If an attack method is blind, the attacker's packets can carry (a) a false IP source address (making it difficult for the victim to find the attacker) and (b) a different address on every packet (making it difficult for the victim to block the attack). If the attacker needs to receive traffic from the victim, the attacker must either (c) reveal its own IP address to the victim (which enables the victim to find the attacker or block the attack by filtering) or (d) provide a false address and also subvert network routing mechanisms to divert the returning packets to the attacker (which makes the attack more complex, more difficult, or more expensive). [R3552] $ block (I) A bit string or bit vector of finite length. (See: bit, block cipher. Compare: byte, word.) Usage: An "N-bit block" contains N bits, which usually are numbered from left to right as 1, 2, 3, ..., N. $ block cipher (I) An encryption algorithm that breaks plain text into fixed-size segments and uses the same key to transform each plaintext segment into a fixed-size segment of cipher text. Examples: AES, Blowfish, DEA, IDEA, RC2, and SKIPJACK. (See: block, mode. Compare: stream cipher.) Tutorial: A block cipher can be adapted to have a different external interface, such as that of a stream cipher, by using a mode of cryptographic operation to package the basic algorithm. (See: CBC, CCM, CFB, CMAC, CTR, DEA, ECB, OFB.) $ Blowfish (N) A symmetric block cipher with variable-length key (32 to 448 bits) designed in 1993 by Bruce Schneier as an unpatented, license-free, royalty-free replacement for DES or IDEA. [Schn] (See: Twofish.) Shirey Informational [Page 38] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ brain-damaged (D) /slang/ "Obviously wrong: extremely poorly designed. Calling something brain-damaged is very extreme. The word implies that the thing is completely unusable, and that its failure to work is due to poor design, not accident." [NCSSG] (See: flaw.) Deprecated Term: It is likely that other cultures use different metaphors for this concept. Therefore, to avoid international misunderstanding, IDOCs SHOULD NOT use this term. (See: Deprecated Usage under "Green Book".) $ brand 1. (I) A distinctive mark or name that identifies a product or business entity. 2. (O) /SET/ The name of a payment card. (See: BCA.) Tutorial: Financial institutions and other companies have founded payment card brands, protect and advertise the brands, establish and enforce rules for use and acceptance of their payment cards, and provide networks to interconnect the financial institutions. These brands combine the roles of issuer and acquirer in interactions with cardholders and merchants. [SET1] $ brand certification authority (BCA) (O) /SET/ A CA owned by a payment card brand, such as MasterCard, Visa, or American Express. [SET2] (See: certification hierarchy, SET.) $ brand CRL identifier (BCI) (O) /SET/ A digitally signed list, issued by a BCA, of the names of CAs for which CRLs need to be processed when verifying signatures in SET messages. [SET2] $ break (I) /cryptography/ To successfully perform cryptanalysis and thus succeed in decrypting data or performing some other cryptographic function, without initially having knowledge of the key that the function requires. (See: penetrate, strength, work factor.) Usage: This term applies to encrypted data or, more generally, to a cryptographic algorithm or cryptographic system. Also, while the most common use is to refer to completely breaking an algorithm, the term is also used when a method is found that substantially reduces the work factor. Shirey Informational [Page 39] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ Brewer-Nash model (N) A security model [BN89] to enforce the Chinese wall policy. (Compare: Bell-LaPadula model, Clark-Wilson model.) Tutorial: All proprietary information in the set of commercial firms F(1), F(2), ..., F(N) is categorized into mutually exclusive conflict-of-interest classes I(1), I(2), ..., I(M) that apply across all firms. Each firm belongs to exactly one class. The Brewer-Nash model has the following mandatory rules: - Brewer-Nash Read Rule: Subject S can read information object O from firm F(i) only if either (a) O is from the same firm as some object previously read by S *or* (b) O belongs to a class I(i) from which S has not previously read any object. (See: object, subject.) - Brewer-Nash Write Rule: Subject S can write information object O to firm F(i) only if (a) S can read O by the Brewer-Nash Read Rule *and* (b) no object can be read by S from a different firm F(j), no matter whether F(j) belongs to the same class as F(i) or to a different class. $ bridge (I) A gateway for traffic flowing at OSIRM Layer 2 between two networks (usually two LANs). (Compare: bridge CA, router.) $ bridge CA (I) A PKI consisting of only a CA that cross-certifies with CAs of some other PKIs. (See: cross-certification. Compare: bridge.) Tutorial: A bridge CA functions as a hub that enables a certificate user in any of the PKIs that attach to the bridge, to validate certificates issued in the other attached PKIs. For example, a bridge CA (BCA) CA1 could cross-certify with four ^ PKIs that have the roots CA1, | CA2, CA3, and CA4. The cross- v certificates that the roots CA2 <-> BCA <-> CA3 exchange with the BCA enable an ^ end entity EE1 certified under | under CA1 in PK1 to construct v a certification path needed to CA4 validate the certificate of end entity EE2 under CA2, CA1 -> BCA -> CA2 -> EE2 or vice versa. CA2 -> BCA -> CA1 -> EE1 Shirey Informational [Page 40] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ British Standard 7799 (N) Part 1 of the standard is a code of practice for how to secure an information system. Part 2 specifies the management framework, objectives, and control requirements for information security management systems. [BS7799] (See: ISO 17799.) $ browser (I) A client computer program that can retrieve and display information from servers on the World Wide Web. Examples: Netscape Navigator and Microsoft Internet Explorer. $ brute force (I) A cryptanalysis technique or other kind of attack method involving an exhaustive procedure that tries a large number of possible solutions to the problem. (See: impossible, strength, work factor.) Tutorial: In some cases, brute force involves trying all of the possibilities. For example, for cipher text where the analyst already knows the decryption algorithm, a brute-force technique for finding matching plain text is to decrypt the message with every possible key. In other cases, brute force involves trying a large number of possibilities but substantially fewer than all of them. For example, given a hash function that produces an N-bit hash result, the probability is greater than 1/2 that the analyst will find two inputs that have the same hash result after trying only 2**(N/2) randomly chosen inputs. (See: birthday attack.) $ BS7799 (N) See: British Standard 7799. $ buffer overflow (I) Any attack technique that exploits a vulnerability resulting from computer software or hardware that does not check for exceeding the bounds of a storage area when data is written into a sequence of storage locations beginning in that area. Tutorial: By causing a normal system operation to write data beyond the bounds of a storage area, the attacker seeks to either disrupt system operation or cause the system to execute malicious software inserted by the attacker. $ buffer zone (I) A neutral internetwork segment used to connect other segments that each operate under a different security policy. Shirey Informational [Page 41] RFC 4949 Internet Security Glossary, Version 2 August 2007 Tutorial: To connect a private network to the Internet or some other relatively public network, one could construct a small, separate, isolated LAN and connect it to both the private network and the public network; one or both of the connections would implement a firewall to limit the traffic that could pass through the buffer zone. $ bulk encryption 1. (I) Encryption of multiple channels by aggregating them into a single transfer path and then encrypting that path. (See: channel.) 2. (O) "Simultaneous encryption of all channels of a multichannel telecommunications link." [C4009] (Compare: bulk keying material.) Usage: The use of "simultaneous" in definition 2 could be interpreted to mean that multiple channels are encrypted separately but at the same time. However, the common meaning of the term is that multiple data flows are combined into a single stream and then that stream is encrypted as a whole. $ bulk key (D) In a few published descriptions of hybrid encryption for SSH, Windows 2000, and other applications, this term refers to a symmetric key that (a) is used to encrypt a relatively large amount of data and (b) is itself encrypted with a public key. (Compare: bulk keying material, session key.) Example: To send a large file to Bob, Alice (a) generates a symmetric key and uses it to encrypt the file (i.e., encrypt the bulk of the information that is to be sent) and then (b) encrypts that symmetric key (the "bulk key") with Bob's public key. Deprecated Term: IDOCs SHOULD NOT use this term or definition; the term is not well-established and could be confused with the established term "bulk keying material". Instead, use "symmetric key" and carefully explain how the key is applied. $ bulk keying material (N) Refers to handling keying material in large quantities, e.g., as a dataset that contains many items of keying material. (See: type 0. Compare: bulk key, bulk encryption.) $ bump-in-the-stack (I) An implementation approach that places a network security mechanism inside the system that is to be protected. (Compare: bump-in-the-wire.) Shirey Informational [Page 42] RFC 4949 Internet Security Glossary, Version 2 August 2007 Example: IPsec can be implemented inboard, in the protocol stack of an existing system or existing system design, by placing a new layer between the existing IP layer and the OSIRM Layer 3 drivers. Source code access for the existing stack is not required, but the system that contains the stack does need to be modified [R4301]. $ bump-in-the-wire (I) An implementation approach that places a network security mechanism outside of the system that is to be protected. (Compare: bump-in-the-stack.) Example: IPsec can be implemented outboard, in a physically separate device, so that the system that receives the IPsec protection does not need to be modified at all [R4301]. Military- grade link encryption has mainly been implemented as bump-in-the- wire devices. $ business-case analysis (N) An extended form of cost-benefit analysis that considers factors beyond financial metrics, including security factors such as the requirement for security services, their technical and programmatic feasibility, their qualitative benefits, and associated risks. (See: risk analysis.) $ byte (I) A fundamental unit of computer storage; the smallest addressable unit in a computer's architecture. Usually holds one character of information and, today, usually means eight bits. (Compare: octet.) Usage: Understood to be larger than a "bit", but smaller than a "word". Although "byte" almost always means "octet" today, some computer architectures have had bytes in other sizes (e.g., six bits, nine bits). Therefore, an STD SHOULD state the number of bits in a byte where the term is first used in the STD. $ C field (D) See: Compartments field. $ C1 or C2 computer system (O) /TCSEC/ See: Tutorial under "Trusted Computer System Evaluation Criteria". $ CA (I) See: certification authority. Shirey Informational [Page 43] RFC 4949 Internet Security Glossary, Version 2 August 2007 $ CA certificate (D) "A [digital] certificate for one CA issued by another CA." [X509] Deprecated Definition: IDOCs SHOULD NOT use the term with this definition; the definition is ambiguous with regard to how the certificate is constructed and how it is intended to be used. IDOCs that use this term SHOULD provide a technical definition for it. (See: certificate profile.) Tutorial: There is no single, obvious choice for a technical definition of this term. Different PKIs can use different certificate profiles, and X.509 provides several choices of how to issue certificates to CAs. For example, one possible definition is the following: A v3 X.509 public-key certificate that has a "basicConstraints" extension containing a "cA" value of "TRUE". That would specifically indicate that "the certified public key may be used to verify certificate signatures", i.e., that the private key may be used by a CA. However, there also are other ways to indicate such usage. The certificate may have a "key Usage" extension that indicates the purposes for which the public key may be used, and one of the values that X.509 defines for that extension is "keyCertSign", to indicate that the certificate may be used for verifying a CA's signature on certificates. If "keyCertSign" is present in a certificate that also has a "basicConstraints" extension, then "cA" is set to "TRUE" in that extension. Alternatively, a CA could be issued a certificate in which "keyCertSign" is asserted without "basicConstraints" being present; and an entity that acts as a CA could be issued a certificate with "keyUsage" set to other values, either with or without "keyCertSign". $ CA domain (N) /PKI/ A security policy domain that "consists of a CA and its subjects [i.e., the entities named in the certificates issued by the CA]. Sometimes referred to as a PKI domain." [PAG] (See: domain.) $ Caesar cipher (I) A cipher that is defined for an alphabet of N characters, A(1), A(2), ..., A(N), and creates cipher text by replacing each plaintext character A(i) by A(i+K, mod N) for some 0