S <- 4. Definitions -> U


T

   $ TACACS
      (I) See: Terminal Access Controller (TAC) Access Control System.
   $ TACACS+
      (I) A TCP-based protocol that improves on TACACS by separating the
      functions of authentication, authorization, and accounting and by
      encrypting all traffic between the network access server and

Shirey                       Informational                    [Page 300]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      authentication server. TACACS+ is extensible to allow any
      authentication mechanism to be used with TACACS+ clients.
   $ tamper
      (I) Make an unauthorized modification in a system that alters the
      system's functioning in a way that degrades the security services
      that the system was intended to provide. (See: QUADRANT. Compare:
      secondary definitions under "corruption" and "misuse".)
   $ tamper-evident
      (I) A characteristic of a system component that provides evidence
      that an attack has been attempted on that component or system.
      Usage: Usually involves physical evidence. (See: tamper.)
   $ tamper-resistant
      (I) A characteristic of a system component that provides passive
      protection against an attack. (See: tamper.)
      Usage: Usually involves physical means of protection.
   $ tampering
      (I) /threat action/ See: secondary definitions under "corruption"
      and "misuse".
   $ target of evaluation (TOE)
      (N) /Common Criteria/ An information technology product or system
      that is the subject of a security evaluation, together with the
      product's associated administrator and user documentation.
      (Compare: protection profile.)
      Tutorial: The security characteristics of the target of evaluation
      (TOE) are described in specific terms by a corresponding security
      target, or in more general terms by a protection profile. In
      Common Criteria philosophy, it is important that a TOE be
      evaluated against the specific set of criteria expressed in the
      target. This evaluation consists of rigorous analysis and testing
      performed by an accredited, independent laboratory. The scope of a
      TOE evaluation is set by the EAL and other requirements specified
      in the target. Part of this process is an evaluation of the target
      itself, to ensure that it is correct, complete, and internally
      consistent and can be used as the baseline for the TOE evaluation.
   $ TCB
      (N) See: trusted computing base.
   $ TCC field
      (I) See: Transmission Control Code field.

Shirey                       Informational                    [Page 301]
RFC 4949         Internet Security Glossary, Version 2       August 2007

   $ TCG
      (N) See: Trusted Computing Group.
   $ TCP
      (I) See: Transmission Control Protocol.
   $ TCP/IP
      (I) Synonym for "Internet Protocol Suite".
   $ TCSEC
      (N) See: Trusted Computer System Evaluation Criteria. (Compare:
      TSEC.)
   $ TDEA
      (I) See: Triple Data Encryption Algorithm.
   $ teardrop attack
      (D) /slang/ A denial-of-service attack that sends improperly
      formed IP packet fragments with the intent of causing the
      destination system to fail.
      Deprecated Term: IDOCs that use this term SHOULD state a
      definition for it because the term is often used imprecisely and
      could easily be misunderstood. (See: Deprecated Usage under "Green
      Book".)
   $ technical non-repudiation
      (I) See: (secondary definition under) non-repudiation.
   $ technical security
      (I) Security mechanisms and procedures that are implemented in and
      executed by computer hardware, firmware, or software to provide
      automated protection for a system. (See: security architecture.
      Compare: administrative security.)
   $ Telecommunications Security Word System (TSEC)
      (O) /U.S. Government/ A terminology for designating
      telecommunication security equipment. (Compare: TCSEC.)
      Tutorial: A TSEC designator has the following parts:
      -  Prefix "TSEC/" for items and systems, or suffix "/TSEC" for
         assemblies. (Often omitted when the context is clear.)
      -  First letter, for function: "C" COMSEC equipment system, "G"
         general purpose, "K" cryptographic, "H" crypto-ancillary, "M"
         manufacturing, "N" noncryptographic, "S" special purpose.
      -  Second letter, for type or purpose: "G" key generation, "I"
         data transmission, "L" literal conversion, "N" signal
         conversion, "O" multipurpose, "P" materials production, "S"

Shirey                       Informational                    [Page 302]
RFC 4949         Internet Security Glossary, Version 2       August 2007
         special purpose, "T" testing or checking, "U" television, "W"
         teletypewriter, "X" facsimile, "Y" speech.
      -  Optional third letter, used only in designations of assemblies,
         for type or purpose: "A" advancing, "B" base or cabinet, "C"
         combining, "D" drawer or panel, "E" strip or chassis, "F" frame
         or rack, "G" key generator, "H" keyboard, "I" translator or
         reader, "J" speech processing, "K" keying or permuting, "L"
         repeater, "M" memory or storage, "O" observation, "P" power
         supply or converter, "R" receiver, "S" synchronizing, "T"
         transmitter, "U" printer, "V" removable COMSEC component, "W"
         logic programmer/programming, "X" special purpose.
      -  Model number, usually two or three digits, assigned
         sequentially within each letter combination (e.g., KG-34, KG-
         84).
      -  Optional suffix letter, used to designate a version. First
         version has no letter, next version has "A" (e.g., KG-84, KG-
         84A), etc.
   $ TELNET
      (I) A TCP-based, Application-Layer, Internet Standard protocol
      (RFC 854) for remote login from one host to another.
   $ TEMPEST
      1. (N) Short name for technology and methods for protecting
      against data compromise due to electromagnetic emanations from
      electrical and electronic equipment. [Army, Russ] (See:
      inspectable space, soft TEMPEST, TEMPEST zone. Compare: QUADRANT)
      2. (O) /U.S. Government/ "Short name referring to investigation,
      study, and control of compromising emanations from IS equipment."
      [C4009]
      Deprecated Usage: IDOCs SHOULD NOT use this term as a synonym for
      "electromagnetic emanations security"; instead, use EMSEC. Also,
      the term is NOT an acronym for Transient Electromagnetic Pulse
      Surveillance Technology.
      Tutorial: The U.S. Federal Government issues security policies
      that (a) state specifications and standards for techniques to
      reduce the strength of emanations from systems and reduce the
      ability of unauthorized parties to receive and make use of
      emanations and (b) state rules for applying those techniques.
      Other nations presumably do the same.
   $ TEMPEST zone
      (O) "Designated area [i.e., a physical volume] within a facility
      where equipment with appropriate TEMPEST characteristics ... may


Shirey                       Informational                    [Page 303]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      be operated." [C4009] (See: emanation security, TEMPEST. Compare:
      control zone, inspectable space.)
      Tutorial: The strength of an electromagnetic signal decreases in
      proportion to the square of the distance between the source and
      the receiver. Therefore, EMSEC for electromagnetic signals can be
      achieved by a combination of (a) reducing the strength of
      emanations to a defined level and (b) establishing around that
      equipment an appropriately sized physical buffer zone from which
      unauthorized entities are excluded. By making the zone large
      enough, it is possible to limit the signal strength available to
      entities outside the zone to a level lower than can be received
      and read with known, state-of-the-art methods. Typically, the need
      for and size of a TEMPEST zone established by a security policy
      depends not only on the measured level of signal emitted by
      equipment, but also on the perceived threat level in the
      equipment's environment.
   $ Terminal Access Controller (TAC) Access Control System (TACACS)
      (I) A UDP-based authentication and access control protocol [R1492]
      in which a network access server receives an identifier and
      password from a remote terminal and passes them to a separate
      authentication server for verification. (See: TACACS+.)
      Tutorial: TACACS can provide service not only for network access
      servers but also routers and other networked computing devices via
      one or more centralized authentication servers. TACACS was
      originally developed for ARPANET and has evolved for use in
      commercial equipment.
   $ TESS
      (I) See: The Exponential Encryption System.
   $ The Exponential Encryption System (TESS)
      (I) A system of separate but cooperating cryptographic mechanisms
      and functions for the secure authenticated exchange of
      cryptographic keys, the generation of digital signatures, and the
      distribution of public keys. TESS uses asymmetric cryptography,
      based on discrete exponentiation, and a structure of self-
      certified public keys. [R1824]
   $ theft
      (I) /threat action/ See: secondary definitions under
      "interception" and "misappropriation".
   $ threat
      1a. (I) A potential for violation of security, which exists when
      there is an entity, circumstance, capability, action, or event

Shirey                       Informational                    [Page 304]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      that could cause harm. (See: dangling threat, INFOCON level,
      threat action, threat agent, threat consequence. Compare: attack,
      vulnerability.)
      1b. (N) Any circumstance or event with the potential to adversely
      affect a system through unauthorized access, destruction,
      disclosure, or modification of data, or denial of service. [C4009]
      (See: sensitive information.)
      Usage: (a) Frequently misused with the meaning of either "threat
      action" or "vulnerability". (b) In some contexts, "threat" is used
      more narrowly to refer only to intelligent threats; for example,
      see definition 2 below. (c) In some contexts, "threat" is used
      more broadly to cover both definition 1 and other concepts, such
      as in definition 3 below.
      Tutorial: A threat is a possible danger that might exploit a
      vulnerability. Thus, a threat may be intentional or not:
      -  "Intentional threat": A possibility of an attack by an
         intelligent entity (e.g., an individual cracker or a criminal
         organization).
      -  "Accidental threat": A possibility of human error or omission,
         unintended equipment malfunction, or natural disaster (e.g.,
         fire, flood, earthquake, windstorm, and other causes listed in
         [FP031]).
      The Common Criteria characterizes a threat in terms of (a) a
      threat agent, (b) a presumed method of attack, (c) any
      vulnerabilities that are the foundation for the attack, and (d)
      the system resource that is attacked. That characterization agrees
      with the definitions in this Glossary (see: diagram under
      "attack").
      2. (O) The technical and operational ability of a hostile entity
      to detect, exploit, or subvert a friendly system and the
      demonstrated, presumed, or inferred intent of that entity to
      conduct such activity.
      Tutorial: To be likely to launch an attack, an adversary must have
      (a) a motive to attack, (b) a method or technical ability to make
      the attack, and (c) an opportunity to appropriately access the
      targeted system.
      3. (D) "An indication of an impending undesirable event." [Park]
      Deprecated Definition: IDOCs SHOULD NOT use this term with
      definition 3 because the definition is ambiguous; the definition
      was intended to include the following three meanings:

Shirey                       Informational                    [Page 305]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      -  "Potential threat": A possible security violation; i.e., the
         same as definition 1.
      -  "Active threat": An expression of intent to violate security.
         (Context usually distinguishes this meaning from the previous
         one.)
      -  "Accomplished threat" or "actualized threat": That is, a threat
         action. Deprecated Usage: IDOCs SHOULD NOT use the term
         "threat" with this meaning; instead, use "threat action".
   $ threat action
      (I) A realization of a threat, i.e., an occurrence in which system
      security is assaulted as the result of either an accidental event
      or an intentional act. (See: attack, threat, threat consequence.)
      Tutorial: A complete security architecture deals with both
      intentional acts (i.e., attacks) and accidental events [FP031].
      (See: various kinds of threat actions defined under the four kinds
      of "threat consequence".)
   $ threat agent
      (I) A system entity that performs a threat action, or an event
      that results in a threat action.
   $ threat analysis
      (I) An analysis of the threat actions that might affect a system,
      primarily emphasizing their probability of occurrence but also
      considering their resulting threat consequences. Example: RFC
      3833. (Compare: risk analysis.)
   $ threat consequence
      (I) A security violation that results from a threat action.
      Tutorial: The four basic types of threat consequence are
      "unauthorized disclosure", "deception", "disruption", and
      "usurpation". (See main Glossary entries of each of these four
      terms for lists of the types of threat actions that can result in
      these consequences.)
   $ thumbprint
      1. (I) A pattern of curves formed by the ridges on the tip of a
      thumb. (See: biometric authentication, fingerprint.)
      2. (D) Synonym for some type of "hash result". (See: biometric
      authentication. Compare: fingerprint.)
      Deprecated Usage: IDOCs SHOULD NOT use this term with definition 2
      because that meaning mixes concepts in a potentially misleading
      way.

Shirey                       Informational                    [Page 306]
RFC 4949         Internet Security Glossary, Version 2       August 2007

   $ ticket
      (I) Synonym for "capability token".
      Tutorial: A ticket is usually granted by a centralized access
      control server (ticket-granting agent) to authorize access to a
      system resource for a limited time. Tickets can be implemented
      with either symmetric cryptography (see: Kerberos) or asymmetric
      cryptography (see: attribute certificate).
   $ tiger team
      (O) A group of evaluators employed by a system's managers to
      perform penetration tests on the system.
      Deprecated Usage: 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".)
   $ time stamp
      1. (I) /noun/ With respect to a data object, a label or marking in
      which is recorded the time (time of day or other instant of
      elapsed time) at which the label or marking was affixed to the
      data object. (See: Time-Stamp Protocol.)
      2. (O) /noun/ "With respect to a recorded network event, a data
      field in which is recorded the time (time of day or other instant
      of elapsed time) at which the event took place." [A1523]
      Tutorial: A time stamp can be used as evidence to prove that a
      data object existed (or that an event occurred) at or before a
      particular time. For example, a time stamp might be used to prove
      that a digital signature based on a private key was created while
      the corresponding public-key certificate was valid, i.e., before
      the certificate either expired or was revoked. Establishing this
      proof would enable the certificate to be used after its expiration
      or revocation, to verify a signature that was created earlier.
      This kind of proof is required as part of implementing PKI
      services, such as non-repudiation service, and long-term security
      services, such as audit.
   $ Time-Stamp Protocol
      (I) An Internet protocol (RFC 3161) that specifies how a client
      requests and receives a time stamp from a server for a data object
      held by the client.
      Tutorial: The protocol describes the format of (a) a request sent
      to a time-stamp authority and (b) the response that is returned
      containing a time stamp. The authority creates the stamp by

Shirey                       Informational                    [Page 307]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      concatenating (a) a hash value of the input data object with (b) a
      UTC time value and other parameters (policy OID, serial number,
      indication of time accuracy, nonce, DN of the authority, and
      various extensions), and then signing that dataset with the
      authority's private key as specified in CMS. Such an authority
      typically would operate as a trusted third-party service, but
      other operational models might be used.
   $ timing channel
      (I) See: covert timing channel.
   $ TKEY
      (I) A mnemonic referring to an Internet protocol (RFC 2930) for
      establishing a shared secret key between a DNS resolver and a DNS
      name server. (See: TSIG.)
   $ TLS
      (I) See: Transport Layer Security.
   $ TLSP
      (N) See: Transport Layer Security Protocol.
   $ TOE
      (N) See: target of evaluation.
   $ token
      1. (I) /cryptography/ See: cryptographic token. (Compare: dongle.)
      2. (I) /access control/ An object that is used to control access
      and is passed between cooperating entities in a protocol that
      synchronizes use of a shared resource. Usually, the entity that
      currently holds the token has exclusive access to the resource.
      (See: capability token.)
      Usage: This term is heavily overloaded in the computing
      literature; therefore, IDOCs SHOULD NOT use this term with any
      definition other than 1 or 2.
      3a. (D) /authentication/ A data object or a physical device used
      to verify an identity in an authentication process.
      3b. (D) /U.S. Government/ Something that the claimant in an
      authentication process (i.e., the entity that claims an identity)
      possesses and controls, and uses to prove the claim during the
      verification step of the process. [SP63]
      Deprecated usage: IDOCs SHOULD NOT use this term with definitions
      3a and 3b; instead, use more specifically descriptive and

Shirey                       Informational                    [Page 308]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      informative terms such as "authentication information" or
      "cryptographic token", depending on what is meant.
      NIST defines four types of claimant tokens for electronic
      authentication in an information system [SP63]. IDOCs SHOULD NOT
      use these four NIST terms; they mix concepts in potentially
      confusing ways and duplicate the meaning of better-established
      terms. These four terms can be avoided by using more specifically
      descriptive terms as follows:
      -  NIST "hard token": A hardware device that contains a protected
         cryptographic key. (This is a type of "cryptographic token",
         and the key is a type of "authentication information".)
      -  NIST "one-time password device token": A personal hardware
         device that generates one-time passwords. (One-time passwords
         are typically generated cryptographically. Therefore, this is a
         type of "cryptographic token", and the key is a type of
         "authentication information".)
      -  NIST "soft token": A cryptographic key that typically is stored
         on disk or some other magnetic media. (The key is a type of
         "authentication information"; "authentication key" would be a
         better description.)
      -  NIST "password token": A secret data value that the claimant
         memorizes. (This is a "password" that is being used as
         "authentication information".)
   $ token backup
      (I) A token management operation that stores sufficient
      information in a database (e.g., in a CAW) to recreate or restore
      a security token (e.g., a smart card) if it is lost or damaged.
   $ token copy
      (I) A token management operation that copies all the personality
      information from one security token to another. However, unlike in
      a token restore operation, the second token is initialized with
      its own, different local security values such as PINs and storage
      keys.
   $ token management
      (I) The process that includes initializing security tokens (e.g.,
      "smart card"), loading data into the tokens, and controlling the
      tokens during their lifecycle. May include performing key
      management and certificate management functions; generating and
      installing PINs; loading user personality data; performing card
      backup, card copy, and card restore operations; and updating
      firmware.




Shirey                       Informational                    [Page 309]
RFC 4949         Internet Security Glossary, Version 2       August 2007

   $ token restore
      (I) A token management operation that loads a security token with
      data for the purpose of recreating (duplicating) the contents
      previously held by that or another token. (See: recovery.)
   $ token storage key
      (I) A cryptographic key used to protect data that is stored on a
      security token.
   $ top CA
      (I) Synonym for "root" in a certification hierarchy. (See: apex
      trust anchor.)
   $ top-level specification
      (I) "A non-procedural description of system behavior at the most
      abstract level; typically a functional specification that omits
      all implementation details." [NCS04] (See: formal top-level
      specification, Tutorial under "security policy".)
      Tutorial: A top-level specification is at a level of abstraction
      below "security model" and above "security architecture" (see:
      Tutorial under "security policy").
      A top-level specification may be descriptive or formal:
      -  "Descriptive top-level specification": One that is written in a
         natural language like English or an informal design notation.
      -  "Formal top-level specification": One that is written in a
         formal mathematical language to enable theorems to be proven
         that show that the specification correctly implements a set of
         formal requirements or a formal security model. (See:
         correctness proof.)
   $ TPM
      (N) See: Trusted Platform Module.
   $ traceback
      (I) Identification of the source of a data packet. (See:
      masquerade, network weaving.)
   $ tracker
      (N) An attack technique for achieving unauthorized disclosure from
      a statistical database. [Denns] (See: Tutorial under "inference
      control".)
   $ traffic analysis
      1. (I) Gaining knowledge of information by inference from
      observable characteristics of a data flow, even if the information
      is not directly available (e.g., when the data is encrypted).

Shirey                       Informational                    [Page 310]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      These characteristics include the identities and locations of the
      source(s) and destination(s) of the flow, and the flow's presence,
      amount, frequency, and duration of occurrence. The object of the
      analysis might be information in SDUs, information in the PCI, or
      both. (See: inference, traffic-flow confidentiality, wiretapping.
      Compare: signal analysis.)
      2. (O) "The inference of information from observation of traffic
      flows (presence, absence, amount, direction, and frequency)."
      [I7498-2]
   $ traffic-flow analysis
      (I) Synonym for "traffic analysis".
   $ traffic-flow confidentiality (TFC)
      1. (I) A data confidentiality service to protect against traffic
      analysis. (See: communications cover.)
      2. (O) "A confidentiality service to protect against traffic
      analysis." [I7498-2]
      Tutorial: Confidentiality concerns involve both direct and
      indirect disclosure of data, and the latter includes traffic
      analysis. However, operational considerations can make TFC
      difficult to achieve. For example, if Alice sends a product idea
      to Bob in an email message, she wants data confidentiality for the
      message's content, and she might also want to conceal the
      destination of the message to hide Bob's identity from her
      competitors. However, the identity of the intended recipient, or
      at least a network address for that recipient, needs to be made
      available to the mail system. Thus, complex forwarding schemes may
      be needed to conceal the ultimate destination as the message
      travels through the open Internet (see: onion routing).
      Later, if Alice uses an ATM during a clandestine visit to
      negotiate with Bob, she might prefer that her bank conceal the
      origin of her transaction, because knowledge of the ATM's location
      might allow a competitor to infer Bob's identity. The bank, on the
      other hand, might prefer to protect only Alice's PIN (see:
      selective-field confidentiality).
      A TFC service can be either full or partial:
      -  "Full TFC": This type of service conceals all traffic
         characteristics.
      -  "Partial TFC": This type of service either (a) conceals some
         but not all of the characteristics or (b) does not completely
         conceal some characteristic.


Shirey                       Informational                    [Page 311]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      On point-to-point data links, full TFC can be provided by
      enciphering all PDUs and also generating a continuous, random data
      stream to seamlessly fill all gaps between PDUs. To a wiretapper,
      the link then appears to be carrying an unbroken stream of
      enciphered data. In other cases -- including on a shared or
      broadcast medium, or end-to-end in a network -- only partial TFC
      is possible, and that may require a combination of techniques. For
      example, a LAN that uses "carrier sense multiple access with
      collision detection" (CSMA/CD; a.k.a. "listen while talk") to
      control access to the medium, relies on detecting intervals of
      silence, which prevents using full TFC. Partial TFC can be
      provided on that LAN by measures such as adding spurious PDUs,
      padding PDUs to a constant size, or enciphering addresses just
      above the Physical Layer; but these measures reduce the efficiency
      with which the LAN can carry traffic. At higher protocol layers,
      SDUs can be protected, but addresses and other items of PCI must
      be visible at the layers below.
   $ traffic key
      (I) A cryptographic key used by a device for protecting
      information that is being transmitted between devices, as opposed
      to protecting information that being is maintained in the device.
      (Compare: storage key.)
   $ traffic padding
      (I) "The generation of spurious instances of communication,
      spurious data units, and/or spurious data within data units."
      [I7498-2]
   $ tranquility property
      (N) /formal model/ Property of a system whereby the security level
      of an object cannot change while the object is being processed by
      the system. (See: Bell-LaPadula model.)
   $ transaction
      1. (I) A unit of interaction between an external entity and a
      system, or between components within a system, that involves a
      series of system actions or events.
      2. (O) "A discrete event between user and systems that supports a
      business or programmatic purpose." [M0404]
      Tutorial: To maintain secure state, transactions need to be
      processed coherently and reliably. Usually, they need to be
      designed to be atomic, consistent, isolated, and durable [Gray]:
      -  "Atomic": All actions and events that comprise the transaction
         are guaranteed to be completed successfully, or else the result
         is as if none at all were executed.

Shirey                       Informational                    [Page 312]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      -  "Consistent": The transaction satisfies correctness constraints
         defined for the data that is being processed.
      -  "Isolated": If two transactions are performed concurrently,
         they do not interfere with each other, and it appears as though
         the system performs one at a time.
      -  "Durable": System state and transaction semantics survive
         system failures.
   $ TRANSEC
      (I) See: transmission security.
   $ Transmission Control Code field (TCC field)
      (I) A data field that provides a means to segregate traffic and
      define controlled communities of interest in the security option
      (option type = 130) of IPv4's datagram header format. The TCC
      values are alphanumeric trigraphs assigned by the U.S. Government
      as specified in RFC 791.
   $ Transmission Control Protocol (TCP)
      (I) An Internet Standard, Transport-Layer protocol (RFC 793) that
      reliably delivers a sequence of datagrams from one computer to
      another in a computer network. (See: TCP/IP.)
      Tutorial: TCP is designed to fit into a layered suite of protocols
      that support internetwork applications. TCP assumes it can obtain
      a simple but potentially unreliable end-to-end datagram service
      (such as IP) from the lower-layer protocols.
   $ transmission security (TRANSEC)
      (I) COMSEC measures that protect communications from interception
      and exploitation by means other than cryptanalysis. Example:
      frequency hopping. (Compare: anti-jam, traffic flow
      confidentiality.)
   $ Transport Layer
      See: Internet Protocol Suite, OSIRM.
   $ Transport Layer Security (TLS)
      (I) TLS is an Internet protocol [R4346] that is based on, and very
      similar to, SSL Version 3.0. (Compare: TLSP.)
      Tutorial: The TLS protocol is misnamed. The name misleadingly
      suggests that TLS is situated in the IPS Transport Layer, but TLS
      is always layered above a reliable Transport-Layer protocol
      (usually TCP) and either layered immediately below or integrated
      with an Application-Layer protocol (often HTTP).



Shirey                       Informational                    [Page 313]
RFC 4949         Internet Security Glossary, Version 2       August 2007

   $ Transport Layer Security Protocol (TLSP)
      (N) An end-to-end encryption protocol (ISO 10736) that provides
      security services at the bottom of OSIRM Layer 4, i.e., directly
      above Layer 3. (Compare: TLS.)
      Tutorial: TLSP evolved directly from SP4.
   $ transport mode
      (I) One of two ways to apply AH or ESP to protect data packets; in
      this mode, the IPsec protocol encapsulates (i.e., the protection
      applies to) the packets of an IPS Transport-Layer protocol (e.g.,
      TCP, UDP), which normally is carried directly above IP in an IPS
      protocol stack. (Compare: tunnel mode.)
      Tutorial: An IPsec transport-mode security association is always
      between two hosts; neither end has the role of a security gateway.
      Whenever either end of an IPsec security association is a security
      gateway, the association is required to be in tunnel mode.
   $ transposition
      (I) /cryptography/ A method of encryption in which elements of the
      plain text retain their original form but undergo some change in
      their sequential position. (Compare: substitution.)
   $ trap door
      (I) Synonym for "back door".
   $ trespass
      (I) /threat action/ See: secondary definition under "intrusion".
   $ Triple Data Encryption Algorithm
      (I) A block cipher that transforms each 64-bit plaintext block by
      applying the DEA three successive times, using either two or three
      different keys for an effective key length of 112 or 168 bits.
      [A9052, SP67]
      Example: A variation proposed for IPsec's ESP uses a 168-bit key,
      consisting of three independent 56-bit values used by the DEA, and
      a 64-bit initialization vector. Each datagram contains an IV to
      ensure that each received datagram can be decrypted even when
      other datagrams are dropped or a sequence of datagrams is
      reordered in transit. [R1851]
   $ triple-wrapped
      (I) /S-MIME/ Data that has been signed with a digital signature,
      then encrypted, and then signed again. [R2634]



Shirey                       Informational                    [Page 314]
RFC 4949         Internet Security Glossary, Version 2       August 2007

   $ Trojan horse
      (I) A computer program that appears to have a useful function, but
      also has a hidden and potentially malicious function that evades
      security mechanisms, sometimes by exploiting legitimate
      authorizations of a system entity that invokes the program. (See:
      malware, spyware. Compare: logic bomb, virus, worm.)
   $ trust
      1. (I) /information system/ A feeling of certainty (sometimes
      based on inconclusive evidence) either (a) that the system will
      not fail or (b) that the system meets its specifications (i.e.,
      the system does what it claims to do and does not perform unwanted
      functions). (See: trust level, trusted system, trustworthy system.
      Compare: assurance.)
      Tutorial: Components of a system can be grouped into three classes
      of trust [Gass]:
      -  "Trusted": The component is responsible for enforcing security
         policy on other components; the system's security depends on
         flawless operation of the component. (See: trusted process.)
      -  "Benign": The component is not responsible for enforcing
         security policy, but it has sensitive authorizations. It must
         be trusted not to intentionally violate security policy, but
         security violations are assumed to be accidental and not likely
         to affect overall system security.
      -  "Untrusted": The component is of unknown or suspicious
         provenance and must be treated as deliberately malicious. (See:
         malicious logic.)
      2. (I) /PKI/ A relationship between a certificate user and a CA in
      which the user acts according to the assumption that the CA
      creates only valid digital certificates.
      Tutorial: "Generally, an entity is said to 'trust' a second entity
      when the first entity makes the assumption that the second entity
      will behave exactly as the first entity expects. This trust may
      apply only for some specific function. The key role of trust in
      [X.509] is to describe the relationship between an entity [i.e., a
      certificate user] and a [CA]; an entity shall be certain that it
      can trust the CA to create only valid and reliable certificates."
      [X509]
   $ trust anchor
      (I) /PKI/ An established point of trust (usually based on the
      authority of some person, office, or organization) from which a
      certificate user begins the validation of a certification path.
      (See: apex trust anchor, path validation, trust anchor CA, trust
      anchor certificate, trust anchor key.)

Shirey                       Informational                    [Page 315]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      Usage: IDOCs that use this term SHOULD state a definition for it
      because it is used in various ways in existing IDOCs and other PKI
      literature. The literature almost always uses this term in a sense
      that is equivalent to this definition, but usage often differs
      with regard to what constitutes the point of trust.
      Tutorial: A trust anchor may be defined as being based on a public
      key, a CA, a public-key certificate, or some combination or
      variation of those:
      -  1. A public key as a point of trust: Although a certification
         path is defined as beginning with a "sequence of public-key
         certificates", an implementation of a path validation process
         might not explicitly handle a root certificate as part of the
         path, but instead begin the process by using a trusted root key
         to verify the signature on a certificate that was issued by the
         root.
         Therefore, "trust anchor" is sometimes defined as just a public
         key. (See: root key, trust anchor key, trusted key.)
      -  2. A CA as a point of trust: A trusted public key is just one
         of the data elements needed for path validation; the IPS path
         validation algorithm [R3280] also needs the name of the CA to
         which that key belongs, i.e., the DN of the issuer of the first
         X.509 certificate to be validated on the path. (See: issue.)
         Therefore, "trust anchor" is sometimes defined as either just a
         CA (where some public key is implied) or as a CA together with
         a specified public key belonging to that CA. (See: root, trust
         anchor CA, trusted CA.)
         Example: "A public key and the name of a [CA] that is used to
         validate the first certificate in a sequence of certificates.
         The trust anchor public key is used to verify the signature on
         a certificate issued by a trust anchor [CA]." [SP57]
      -  3. A public-key certificate as a point of trust: Besides the
         trusted CA's public key and name, the path validation algorithm
         needs to know the digital signature algorithm and any
         associated parameters with which the public key is used, and
         also any constraints that have been placed on the set of paths
         that may be validated using the key. All of this information is
         available from a CA's public-key certificate.
         Therefore, "trust anchor" is sometimes defined as a public-key
         certificate of a CA. (See: root certificate, trust anchor
         certificate, trusted certificate.)

Shirey                       Informational                    [Page 316]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      -  4. Combinations: Combinations and variations of the first three
         definitions are also used in the PKI literature.
         Example: "trust anchor information". The IPS standard for path
         validation [R3280] specifies the information that describes "a
         CA that serves as a trust anchor for the certification path.
         The trust anchor information includes: (a) the trusted issuer
         name, (b) the trusted public key algorithm, (c) the trusted
         public key, and (d) optionally, the trusted public key
         parameters associated with the public key. The trust anchor
         information may be provided to the path processing procedure in
         the form of a self-signed certificate. The trusted anchor
         information is trusted because it was delivered to the path
         processing procedure by some trustworthy out-of-band procedure.
         If the trusted public key algorithm requires parameters, then
         the parameters are provided along with the trusted public key."
   $ trust anchor CA
      (I) A CA that is the subject of a trust anchor certificate or
      otherwise establishes a trust anchor key. (See: root, trusted CA.)
      Tutorial: The selection of a CA to be a trust anchor is a matter
      of policy. Some of the possible choices include (a) the top CA in
      a hierarchical PKI, (b) the CA that issued the verifier's own
      certificate, or (c) any other CA in a network PKI. Different
      applications may rely on different trust anchors, or may accept
      paths that begin with any of a set of trust anchors. The IPS path
      validation algorithm is the same, regardless of the choice.
   $ trust anchor certificate
      (I) A public-key certificate that is used to provide the first
      public key in a certification path. (See: root certificate, trust
      anchor, trusted certificate.)
   $ trust anchor key
      (I) A public key that is used as the first public key in a
      certification path. (See: root key, trust anchor, trusted public
      key.)
   $ trust anchor information
      (I) See: secondary definition under "trust anchor".
   $ trust chain
      (D) Synonym for "certification path". (See: trust anchor, trusted
      certificate.)




Shirey                       Informational                    [Page 317]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      Deprecated Term: IDOCs SHOULD NOT use this term, because it
      unnecessarily duplicates the meaning of the internationally
      standardized term.
      Also, the term mixes concepts in a potentially misleading way.
      Having "trust" involves factors unrelated to simply verifying
      signatures and performing other tests as specified by a standard
      algorithm for path validation (e.g., RFC 3280). Thus, even if a
      user is able to validate a certification path algorithmically, the
      user still might distrust one of the CAs that issued certificates
      in that path or distrust some other aspects of the PKI.
   $ trust-file PKI
      (I) A non-hierarchical PKI in which each certificate user has its
      own local file (which is used by application software) of trust
      anchors, i.e., either public keys or public-key certificates that
      the user trusts as starting points for certification paths. (See:
      trust anchor, web of trust. Compare: hierarchical PKI, mesh PKI.)
      Example: Popular browsers are distributed with an initial file of
      trust anchor certificates, which often are self-signed
      certificates. Users can add certificates to the file or delete
      from it. The file may be directly managed by the user, or the
      user's organization may manage it from a centralized server.
   $ trust hierarchy
      (D) Synonym for "certification hierarchy".
      Deprecated Usage: IDOCs SHOULD NOT use this term because it mixes
      concepts in a potentially misleading way, and because a trust
      hierarchy could be implemented in other ways. (See: trust, trust
      chain, web of trust.)
   $ trust level
      (N) A characterization of a standard of security protection to be
      met by an information system. (See: Common Criteria, TCSEC.)
      Tutorial: A trust level is based not only on (a) the presence of
      security mechanisms, but also on the use of (b) systems
      engineering discipline to properly structure the system and (c)
      implementation analysis to ensure that the system provides an
      appropriate degree of trust.
   $ trusted
      (I) See: secondary definition under "trust".




Shirey                       Informational                    [Page 318]
RFC 4949         Internet Security Glossary, Version 2       August 2007

   $ trusted CA
      (I) A CA upon which a certificate user relies as issuing valid
      certificates; especially a CA that is used as a trust anchor CA.
      (See: certification path, root, trust anchor CA, validation.)
      Tutorial. This trust is transitive to the extent that the X.509
      certificate extensions permit; that is, if a trusted CA issues a
      certificate to another CA, a user that trusts the first CA also
      trusts the second CA if the user succeeds in validating the
      certificate path (see: path validation).
   $ trusted certificate
      (I) A digital certificate that a certificate user accepts as being
      valid "a priori", i.e., without testing the certificate to
      validate it as the final certificate on a certification path;
      especially a certificate that is used as a trust anchor
      certificate. (See: certification path, root certificate, trust
      anchor certificate, trust-file PKI, validation.)
      Tutorial: The acceptance of a certificate as trusted is a matter
      of policy and choice. Usually, a certificate is accepted as
      trusted because the user obtained it by reliable, out-of-band
      means that cause the user to believe the certificate accurately
      binds its subject's name to the subject's public key or other
      attribute values. Many choices are possible; e.g., a trusted
      public-key certificate might be (a) the root certificate in a
      hierarchical PKI, (b) the certificate of the CA that issued the
      user's own certificate in a mesh PKI, or (c) a certificate
      provided with an application that uses a trust-file PKI.
   $ Trusted Computer System Evaluation Criteria (TCSEC)
      (N) A standard for evaluating the security provided by operating
      systems [CSC1, DoD1]. Known as the "Orange Book" because of the
      color of its cover; first document in the Rainbow Series. (See:
      Common Criteria, Deprecated Usage under "Green Book", Orange Book,
      trust level, trusted system. Compare: TSEC.)
      Tutorial: The TCSEC defines classes of hierarchically ordered
      assurance levels for rating computer systems. From highest to
      lowest, the classes are as follows:
      -  Division A:  Verified protection.
           Beyond A1    Beyond current technology. (See: beyond A1.)
           Class  A1    Verified design. (See: SCOMP.)
      -  Division B:  Mandatory protection.
           Class  B3    Security domains.
           Class  B2    Structured protection. (See: Multics.)
           Class  B1    Labeled security protection.


Shirey                       Informational                    [Page 319]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      -  Division C:  Discretionary protection.
           Class  C2    Controlled access protection.
           Class  C1    Discretionary security protection.
      -  Division D:  Minimal protection, i.e., has been evaluated but
         does not meet the requirements for a higher evaluation class.
   $ trusted computing base (TCB)
      (N) "The totality of protection mechanisms within a computer
      system, including hardware, firmware, and software, the
      combination of which is responsible for enforcing a security
      policy." [NCS04] (See: "trusted" under "trust". Compare: TPM.)
   $ Trusted Computing Group (TCG)
      (N) A not-for-profit, industry standards organization formed to
      develop, define, and promote open standards for hardware-enabled
      trusted computing and security technologies, including hardware
      building blocks and software interfaces, across multiple
      platforms, peripherals, and devices. (See: TPM, trusted system.
      Compare: TSIG.)
   $ trusted distribution
      (I) /COMPUSEC/ "A trusted method for distributing the TCB
      hardware, software, and firmware components, both originals and
      updates, that provides methods for protecting the TCB from
      modification during distribution and for detection of any changes
      to the TCB that may occur." [NCS04] (See: code signing,
      configuration control.)
   $ trusted key
      (D) Abbreviation for "trusted public key" and also for other types
      of keys. (See: root key, trust anchor key.)
      Deprecated Usage: IDOCs SHOULD either (a) state a definition for
      this term or (b) use a different, less ambiguous term. This term
      is ambiguous when it stands alone; e.g., it could refer to a
      trusted public key or to a private key or symmetric key that is
      believed to be secure (i.e., not compromised).
   $ trusted path
      1a. (I) /COMPUSEC/ A mechanism by which a computer system user can
      communicate directly and reliably with the TCB and that can only
      be activated by the user or the TCB and cannot be imitated by
      untrusted software within the computer. [NCS04]
      1b. (I) /COMSEC/ A mechanism by which a person or process can
      communicate directly with a cryptographic module and that can only
      be activated by the person, process, or module, and cannot be
      imitated by untrusted software within the module. [FP140]

Shirey                       Informational                    [Page 320]
RFC 4949         Internet Security Glossary, Version 2       August 2007

   $ Trusted Platform Module (TPM)
      (N) The name of a specification, published by the TCG, for a
      microcontroller that can store secured information; and also the
      general name of implementations of that specification. (Compare:
      TCB.)
   $ trusted process
      (I) A system component that has privileges that enable it to
      affect the state of system security and that can, therefore,
      through incorrect or malicious execution, violate the system's
      security policy. (See: privileged process, trusted system.)
   $ trusted public key
      (I) A public key upon which a user relies; especially a public key
      that is used as a trust anchor key. (See: certification path, root
      key, trust anchor key, validation.)
      Tutorial: A trusted public key could be (a) the root key in a
      hierarchical PKI, (b) the key of the CA that issued the user's own
      certificate in a mesh PKI, or (c) any key accepted by the user in
      a trust-file PKI.
   $ trusted recovery
      (I) A process that, after a system has experienced a failure or an
      attack, restores the system to normal operation (or to a secure
      state) without causing a security compromise. (See: recovery.)
   $ trusted subnetwork
      (I) A subnetwork containing hosts and routers that trust each
      other not to engage in active or passive attacks. (There also is
      an assumption that the underlying communication channels, such as
      telephone lines or a LAN, are protected from attack.)
   $ trusted system
      1. (I) /information system/ A system that operates as expected,
      according to design and policy, doing what is required -- despite
      environmental disruption, human user and operator errors, and
      attacks by hostile parties -- and not doing other things [NRC98].
      (See: trust level, trusted process. Compare: trustworthy.)
      2. (N) /multilevel secure/ "A [trusted system is a] system that
      employs sufficient hardware and software assurance measures to
      allow its use for simultaneous processing of a range of sensitive
      or classified information." [NCS04] (See: multilevel security
      mode.)




Shirey                       Informational                    [Page 321]
RFC 4949         Internet Security Glossary, Version 2       August 2007

   $ Trusted Systems Interoperability Group (TSIG)
      (N) A forum of computer vendors, system integrators, and users
      devoted to promoting interoperability of trusted computer systems.
      (See: trusted system. Compare: TCG.)
   $ trustworthy system
      1. (I) A system that not only is trusted, but also warrants that
      trust because the system's behavior can be validated in some
      convincing way, such as through formal analysis or code review.
      (See: trust. Compare: trusted.)
      2. (O) /Digital Signature Guidelines/ "Computer hardware,
      software, and procedures that: (a) are reasonably secure from
      intrusion and misuse; (b) provide a reasonably reliable level of
      availability, reliability, and correct operation; (c) are
      reasonably suited to performing their intended functions; and (d)
      adhere to generally accepted security principles." [DSG]
   $ TSEC
      (O) See: Telecommunications Security Nomenclature System.
      (Compare: TCSEC.)
   $ TSIG
      1. (N) See: Trusted System Interoperability Group.
      2. (I) A mnemonic (presumed to be derived from "Transaction
      SIGnature") referring to an Internet protocol (RFC 2845) for data
      origin authentication and data integrity for certain DNS
      operations. (See: TKEY.)
   $ tunnel
      1. (I) A communication channel created in a computer network by
      encapsulating (i.e., layering) a communication protocol's data
      packets in (i.e., above) a second protocol that normally would be
      carried above, or at the same layer as, the first one. (See: L2TP,
      tunnel mode, VPN. Compare: covert channel.)
      Tutorial: Tunneling can involve almost any two IPS protocol
      layers. For example, a TCP connection between two hosts could
      conceivably be carried above SMTP (i.e., in SMTP messages) as a
      covert channel to evade access controls that a security gateway
      applies to the normal TCP layer that is below SMTP.
      Usually, however, a tunnel is a logical point-to-point link --
      i.e., an OSIRM Layer 2 connection -- created by encapsulating the
      Layer 2 protocol in one of the following three types of IPS
      protocols: (a) an IPS Transport-Layer protocol (such as TCP), (b)
      an IPS Network-Layer or Internet-Layer protocol (such as IP), or

Shirey                       Informational                    [Page 322]
RFC 4949         Internet Security Glossary, Version 2       August 2007

      (c) another Layer 2 protocol. In many cases, the encapsulation is
      accomplished with an extra, intermediate protocol (i.e., a
      "tunneling protocol"; e.g., L2TP) that is layered below the
      tunneled Layer 2 protocol and above the encapsulating protocol.
      Tunneling can be used to move data between computers that use a
      protocol not supported by the network connecting them. Tunneling
      also can enable a computer network to use the services of a second
      network as though the second network were a set of point-to-point
      links between the first network's nodes. (See: VPN.)
      2. (O) /SET/ The name of a SET private extension that indicates
      whether the CA or the payment gateway supports passing encrypted
      messages to the cardholder through the merchant. If so, the
      extension lists OIDs of symmetric encryption algorithms that are
      supported.
   $ tunnel mode
      (I) One of two ways to apply the IPsec protocols (AH and ESP) to
      protect data packets; in this mode, the IPsec protocol
      encapsulates (i.e., the protection applies to) IP packets, rather
      than the packets of higher-layer protocols. (See: tunnel. Compare:
      transport mode.)
      Tutorial: Each end of a tunnel-mode security association may be
      either a host or a security gateway. Whenever either end of an
      IPsec security association is a security gateway, the association
      is required to be in tunnel mode.
   $ two-person control
      (I) The close surveillance and control of a system, a process, or
      materials (especially with regard to cryptography) at all times by
      a minimum of two appropriately authorized persons, each capable of
      detecting incorrect and unauthorized procedures with respect to
      the tasks to be performed and each familiar with established
      security requirements. (See: dual control, no-lone zone.)
   $ Twofish
      (O) A symmetric, 128-bit block cipher with variable key length
      (128, 192, or 256 bits), developed by Counterpane Labs as a
      candidate for the AES. (See: Blowfish.)
   $ type 0 product
      (O) /cryptography, U.S. Government/ Classified cryptographic
      equipment endorsed by NSA for use (when appropriately keyed) in
      electronically distributing bulk keying material.



Shirey                       Informational                    [Page 323]
RFC 4949         Internet Security Glossary, Version 2       August 2007

   $ type 1 key
      (O) /cryptography, U.S. Government/ "Generated and distributed
      under the auspices of NSA for use in a cryptographic device for
      the protection of classified and sensitive national security
      information." [C4009]
   $ type 1 product
      (O) /cryptography, U.S. Government/ "Cryptographic equipment,
      assembly or component classified or certified by NSA for
      encrypting and decrypting classified and sensitive national
      security information when appropriately keyed. Developed using
      established NSA business processes and containing NSA approved
      algorithms. Used to protect systems requiring the most stringent
      protection mechanisms." [C4009]
      Tutorial: The current definition of this term is less specific
      than an earlier version: "Classified or controlled cryptographic
      item endorsed by the NSA for securing classified and sensitive
      U.S. Government information, when appropriately keyed. The term
      refers only to products, and not to information, key, services, or
      controls. Type 1 products contain classified NSA algorithms. They
      are available to U.S. Government users, their contractors, and
      federally sponsored non-U.S. Government activities subject to
      export restrictions in accordance with International Traffic in
      Arms Regulation." [from an earlier version of C4009] (See: ITAR.)
   $ type 2 key
      (O) /cryptography, U.S. Government/ "Generated and distributed
      under the auspices of NSA for use in a cryptographic device for
      the protection of unclassified national security information."
      [C4009]
   $ type 2 product
      (O) /cryptography, U.S. Government/ "Cryptographic equipment,
      assembly, or component certified by NSA for encrypting or
      decrypting sensitive national security information when
      appropriately keyed. Developed using established NSA business
      processes and containing NSA approved algorithms. Used to protect
      systems requiring protection mechanisms exceeding best commercial
      practices including systems used for the protection of
      unclassified national security information." [C4009]
      Tutorial: The current definition of this term is less specific
      than an earlier version: "Unclassified cryptographic equipment,
      assembly, or component, endorsed by the NSA, for use in national
      security systems as defined in Title 40 U.S.C. Section 1452."
      [from an earlier version of C4009] (See: national security system.
      Compare: EUCI.)

Shirey                       Informational                    [Page 324]
RFC 4949         Internet Security Glossary, Version 2       August 2007

   $ type 3 key
      (O) /cryptography, U.S. Government/ "Used in a cryptographic
      device for the protection of unclassified sensitive information,
      even if used in a Type 1 or Type 2 product." [C4009]
   $ type 3 product
      (O) /cryptography, U.S. Government/ "Unclassified cryptographic
      equipment, assembly, or component used, when appropriately keyed,
      for encrypting or decrypting unclassified sensitive U.S.
      Government or commercial information, and to protect systems
      requiring protection mechanisms consistent with standard
      commercial practices. Developed using established commercial
      standards and containing NIST approved cryptographic
      algorithms/modules or successfully evaluated by the National
      Information Assurance Partnership (NIAP)." [C4009]
    type 4 key
      (O) /cryptography, U.S. Government/ "Used by a cryptographic
      device in support of its Type 4 functionality; i.e., any provision
      of key that lacks U.S. Government endorsement or oversight."
      [C4009]
   $ type 4 product
      (O) /cryptography, U.S. Government/ "Unevaluated commercial
      cryptographic equipment, assemblies, or components that neither
      NSA nor NIST certify for any Government usage. These products are
      typically delivered as part of commercial offerings and are
      commensurate with the vendor's commercial practices. These
      products may contain either vendor proprietary algorithms,
      algorithms registered by NIST, or algorithms registered by NIST
      and published in a FIPS." [C4009]

S <- 4. Definitions -> U