Internet Engineering Task Force (IETF)                       E. Rescorla
Request for Comments: 8827                                       Mozilla
Category: Standards Track                                   January 2021
ISSN: 2070-1721


                     WebRTC Security Architecture

Abstract

  This document defines the security architecture for WebRTC, a
  protocol suite intended for use with real-time applications that can
  be deployed in browsers -- "real-time communication on the Web".

Status of This Memo

  This is an Internet Standards Track document.

  This document is a product of the Internet Engineering Task Force
  (IETF).  It represents the consensus of the IETF community.  It has
  received public review and has been approved for publication by the
  Internet Engineering Steering Group (IESG).  Further information on
  Internet Standards is available in Section 2 of RFC 7841.

  Information about the current status of this document, any errata,
  and how to provide feedback on it may be obtained at
  https://www.rfc-editor.org/info/rfc8827.

Copyright Notice

  Copyright (c) 2021 IETF Trust and the persons identified as the
  document authors.  All rights reserved.

  This document is subject to BCP 78 and the IETF Trust's Legal
  Provisions Relating to IETF Documents
  (https://trustee.ietf.org/license-info) in effect on the date of
  publication of this document.  Please review these documents
  carefully, as they describe your rights and restrictions with respect
  to this document.  Code Components extracted from this document must
  include Simplified BSD License text as described in Section 4.e of
  the Trust Legal Provisions and are provided without warranty as
  described in the Simplified BSD License.

  This document may contain material from IETF Documents or IETF
  Contributions published or made publicly available before November
  10, 2008.  The person(s) controlling the copyright in some of this
  material may not have granted the IETF Trust the right to allow
  modifications of such material outside the IETF Standards Process.
  Without obtaining an adequate license from the person(s) controlling
  the copyright in such materials, this document may not be modified
  outside the IETF Standards Process, and derivative works of it may
  not be created outside the IETF Standards Process, except to format
  it for publication as an RFC or to translate it into languages other
  than English.

Table of Contents

  1.  Introduction
  2.  Terminology
  3.  Trust Model
    3.1.  Authenticated Entities
    3.2.  Unauthenticated Entities
  4.  Overview
    4.1.  Initial Signaling
    4.2.  Media Consent Verification
    4.3.  DTLS Handshake
    4.4.  Communications and Consent Freshness
  5.  SDP Identity Attribute
    5.1.  Offer/Answer Considerations
      5.1.1.  Generating the Initial SDP Offer
      5.1.2.  Generating an SDP Answer
      5.1.3.  Processing an SDP Offer or Answer
      5.1.4.  Modifying the Session
  6.  Detailed Technical Description
    6.1.  Origin and Web Security Issues
    6.2.  Device Permissions Model
    6.3.  Communications Consent
    6.4.  IP Location Privacy
    6.5.  Communications Security
  7.  Web-Based Peer Authentication
    7.1.  Trust Relationships: IdPs, APs, and RPs
    7.2.  Overview of Operation
    7.3.  Items for Standardization
    7.4.  Binding Identity Assertions to JSEP Offer/Answer
          Transactions
      7.4.1.  Carrying Identity Assertions
    7.5.  Determining the IdP URI
      7.5.1.  Authenticating Party
      7.5.2.  Relying Party
    7.6.  Requesting Assertions
    7.7.  Managing User Login
  8.  Verifying Assertions
    8.1.  Identity Formats
  9.  Security Considerations
    9.1.  Communications Security
    9.2.  Privacy
    9.3.  Denial of Service
    9.4.  IdP Authentication Mechanism
      9.4.1.  PeerConnection Origin Check
      9.4.2.  IdP Well-Known URI
      9.4.3.  Privacy of IdP-Generated Identities and the Hosting
              Site
      9.4.4.  Security of Third-Party IdPs
        9.4.4.1.  Confusable Characters
      9.4.5.  Web Security Feature Interactions
        9.4.5.1.  Popup Blocking
        9.4.5.2.  Third Party Cookies
  10. IANA Considerations
  11. References
    11.1.  Normative References
    11.2.  Informative References
  Acknowledgements
  Author's Address

1.  Introduction

  The Real-Time Communications on the Web (RTCWEB) Working Group
  standardized protocols for real-time communications between Web
  browsers, generally called "WebRTC" [RFC8825].  The major use cases
  for WebRTC technology are real-time audio and/or video calls, Web
  conferencing, and direct data transfer.  Unlike most conventional
  real-time systems (e.g., SIP-based [RFC3261] soft phones), WebRTC
  communications are directly controlled by some Web server, via a
  JavaScript (JS) API as shown in Figure 1.

                           +----------------+
                           |                |
                           |   Web Server   |
                           |                |
                           +----------------+
                               ^        ^
                              /          \
                      HTTP   /            \   HTTP
                            /              \
                           /                \
                          v                  v
                       JS API              JS API
                 +-----------+            +-----------+
                 |           |    Media   |           |
                 |  Browser  |<---------->|  Browser  |
                 |           |            |           |
                 +-----------+            +-----------+

                     Figure 1: A Simple WebRTC System

  A more complicated system might allow for inter-domain calling, as
  shown in Figure 2.  The protocol to be used between the domains is
  not standardized by WebRTC, but given the installed base and the form
  of the WebRTC API is likely to be something SDP-based like SIP or
  something like the Extensible Messaging and Presence Protocol (XMPP)
  [RFC6120].

            +--------------+                +--------------+
            |              | SIP, XMPP, ... |              |
            |  Web Server  |<-------------->|  Web Server  |
            |              |                |              |
            +--------------+                +--------------+
                   ^                                ^
                   |                                |
             HTTP  |                                |  HTTP
                   |                                |
                   v                                v
                   JS API                       JS API
             +-----------+                     +-----------+
             |           |        Media        |           |
             |  Browser  |<------------------->|  Browser  |
             |           |                     |           |
             +-----------+                     +-----------+

                  Figure 2: A Multidomain WebRTC System

  This system presents a number of new security challenges, which are
  analyzed in [RFC8826].  This document describes a security
  architecture for WebRTC which addresses the threats and requirements
  described in that document.

2.  Terminology

  The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
  "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
  "OPTIONAL" in this document are to be interpreted as described in
  BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all
  capitals, as shown here.

3.  Trust Model

  The basic assumption of this architecture is that network resources
  exist in a hierarchy of trust, rooted in the browser, which serves as
  the user's Trusted Computing Base (TCB).  Any security property which
  the user wishes to have enforced must be ultimately guaranteed by the
  browser (or transitively by some property the browser verifies).
  Conversely, if the browser is compromised, then no security
  guarantees are possible.  Note that there are cases (e.g., Internet
  kiosks) where the user can't really trust the browser that much.  In
  these cases, the level of security provided is limited by how much
  they trust the browser.

  Optimally, we would not rely on trust in any entities other than the
  browser.  However, this is unfortunately not possible if we wish to
  have a functional system.  Other network elements fall into two
  categories: those which can be authenticated by the browser and thus
  can be granted permissions to access sensitive resources, and those
  which cannot be authenticated and thus are untrusted.

3.1.  Authenticated Entities

  There are two major classes of authenticated entities in the system:

  Calling services:  Web sites whose origin we can verify (optimally
     via HTTPS, but in some cases because we are on a topologically
     restricted network, such as behind a firewall, and can infer
     authentication from firewall behavior).

  Other users:  WebRTC peers whose origin we can verify
     cryptographically (optimally via DTLS-SRTP).

  Note that merely being authenticated does not make these entities
  trusted.  For instance, just because we can verify that
  <https://www.example.org/> is owned by Dr. Evil does not mean that we
  can trust Dr. Evil to access our camera and microphone.  However, it
  gives the user an opportunity to determine whether they wish to trust
  Dr. Evil or not; after all, if they desire to contact Dr. Evil
  (perhaps to arrange for ransom payment), it's safe to temporarily
  give them access to the camera and microphone for the purpose of the
  call, but they don't want Dr. Evil to be able to access their camera
  and microphone other than during the call.  The point here is that we
  must first identify other elements before we can determine whether
  and how much to trust them.  Additionally, sometimes we need to
  identify the communicating peer before we know what policies to
  apply.

3.2.  Unauthenticated Entities

  Other than the above entities, we are not generally able to identify
  other network elements; thus, we cannot trust them.  This does not
  mean that it is not possible to have any interaction with them, but
  it means that we must assume that they will behave maliciously and
  design a system which is secure even if they do so.

4.  Overview

  This section describes a typical WebRTC session and shows how the
  various security elements interact and what guarantees are provided
  to the user.  The example in this section is a "best case" scenario
  in which we provide the maximal amount of user authentication and
  media privacy with the minimal level of trust in the calling service.
  Simpler versions with lower levels of security are also possible and
  are noted in the text where applicable.  It's also important to
  recognize the tension between security (or performance) and privacy.
  The example shown here is aimed towards settings where we are more
  concerned about secure calling than about privacy, but as we shall
  see, there are settings where one might wish to make different
  tradeoffs -- this architecture is still compatible with those
  settings.

  For the purposes of this example, we assume the topology shown in the
  figures below.  This topology is derived from the topology shown in
  Figure 1, but separates Alice's and Bob's identities from the process
  of signaling.  Specifically, Alice and Bob have relationships with
  some Identity Provider (IdP) that supports a protocol (such as OpenID
  Connect) that can be used to demonstrate their identity to other
  parties.  For instance, Alice might have an account with a social
  network which she can then use to authenticate to other Web sites
  without explicitly having an account with those sites; this is a
  fairly conventional pattern on the Web. Section 7.1 provides an
  overview of IdPs and the relevant terminology.  Alice and Bob might
  have relationships with different IdPs as well.  Note: The IdP
  mechanism described here has not seen wide adoption.  See Section 7
  for more on the status of IdP-based authentication.

  This separation of identity provision and signaling isn't
  particularly important in "closed world" cases where Alice and Bob
  are users on the same social network and have identities based on
  that domain (Figure 3).  However, there are important settings where
  that is not the case, such as federation (calls from one domain to
  another; see Figure 4) and calling on untrusted sites, such as where
  two users who have a relationship via a given social network want to
  call each other on another, untrusted, site, such as a poker site.

  Note that the servers themselves are also authenticated by an
  external identity service, the SSL/TLS certificate infrastructure
  (not shown).  As is conventional in the Web, all identities are
  ultimately rooted in that system.  For instance, when an IdP makes an
  identity assertion, the Relying Party consuming that assertion is
  able to verify because it is able to connect to the IdP via HTTPS.

                              +----------------+
                              |                |
                              |     Signaling  |
                              |     Server     |
                              |                |
                              +----------------+
                                  ^        ^
                                 /          \
                         HTTPS  /            \   HTTPS
                               /              \
                              /                \
                             v                  v
                          JS API              JS API
                    +-----------+            +-----------+
                    |           |    Media   |           |
              Alice |  Browser  |<---------->|  Browser  | Bob
                    |           | (DTLS+SRTP)|           |
                    +-----------+            +-----------+
                          ^      ^--+     +--^     ^
                          |         |     |        |
                          v         |     |        v
                    +-----------+   |     |  +-----------+
                    |           |<--------+  |           |
                    |   IdP1    |   |        |    IdP2   |
                    |           |   +------->|           |
                    +-----------+            +-----------+

                 Figure 3: A Call with IdP-Based Identity

  Figure 4 shows essentially the same calling scenario but with a call
  between two separate domains (i.e., a federated case), as in
  Figure 2.  As mentioned above, the domains communicate by some
  unspecified protocol, and providing separate signaling and identity
  allows for calls to be authenticated regardless of the details of the
  inter-domain protocol.

          +----------------+    Unspecified    +----------------+
          |                |      protocol     |                |
          |    Signaling   |<----------------->|    Signaling   |
          |    Server      |  (SIP, XMPP, ...) |    Server      |
          |                |                   |                |
          +----------------+                   +----------------+
                   ^                                   ^
                   |                                   |
             HTTPS |                                   | HTTPS
                   |                                   |
                   |                                   |
                   v                                   v
                JS API                               JS API
          +-----------+                             +-----------+
          |           |             Media           |           |
    Alice |  Browser  |<--------------------------->|  Browser  | Bob
          |           |           DTLS+SRTP         |           |
          +-----------+                             +-----------+
                ^      ^--+                      +--^     ^
                |         |                      |        |
                v         |                      |        v
          +-----------+   |                      |  +-----------+
          |           |<-------------------------+  |           |
          |   IdP1    |   |                         |    IdP2   |
          |           |   +------------------------>|           |
          +-----------+                             +-----------+

            Figure 4: A Federated Call with IdP-Based Identity

4.1.  Initial Signaling

  For simplicity, assume the topology in Figure 3.  Alice and Bob are
  both users of a common calling service; they both have approved the
  calling service to make calls (we defer the discussion of device
  access permissions until later).  They are both connected to the
  calling service via HTTPS and so know the origin with some level of
  confidence.  They also have accounts with some IdP.  This sort of
  identity service is becoming increasingly common in the Web
  environment (with technologies such as Federated Google Login,
  Facebook Connect, OAuth, OpenID, WebFinger), and is often provided as
  a side effect service of a user's ordinary accounts with some
  service.  In this example, we show Alice and Bob using a separate
  identity service, though the identity service may be the same entity
  as the calling service or there may be no identity service at all.

  Alice is logged onto the calling service and decides to call Bob. She
  can see from the calling service that he is online and the calling
  service presents a JS UI in the form of a button next to Bob's name
  which says "Call".  Alice clicks the button, which initiates a JS
  callback that instantiates a PeerConnection object.  This does not
  require a security check: JS from any origin is allowed to get this
  far.

  Once the PeerConnection is created, the calling service JS needs to
  set up some media.  Because this is an audio/video call, it creates a
  MediaStream with two MediaStreamTracks, one connected to an audio
  input and one connected to a video input.  At this point, the first
  security check is required: untrusted origins are not allowed to
  access the camera and microphone, so the browser prompts Alice for
  permission.

  In the current W3C API, once some streams have been added, Alice's
  browser + JS generates a signaling message [RFC8829] containing:

  *  Media channel information

  *  Interactive Connectivity Establishment (ICE) [RFC8445] candidates

  *  A "fingerprint" attribute binding the communication to a key pair
     [RFC5763].  Note that this key may simply be ephemerally generated
     for this call or specific to this domain, and Alice may have a
     large number of such keys.

  Prior to sending out the signaling message, the PeerConnection code
  contacts the identity service and obtains an assertion binding
  Alice's identity to her fingerprint.  The exact details depend on the
  identity service (though as discussed in Section 7 PeerConnection can
  be agnostic to them), but for now it's easiest to think of as an
  OAuth token.  The assertion may bind other information to the
  identity besides the fingerprint, but at minimum it needs to bind the
  fingerprint.

  This message is sent to the signaling server, e.g., by fetch()
  [fetch] or by WebSockets [RFC6455], over TLS [RFC8446].  The
  signaling server processes the message from Alice's browser,
  determines that this is a call to Bob, and sends a signaling message
  to Bob's browser (again, the format is currently undefined).  The JS
  on Bob's browser processes it, and alerts Bob to the incoming call
  and to Alice's identity.  In this case, Alice has provided an
  identity assertion and so Bob's browser contacts Alice's IdP (again,
  this is done in a generic way so the browser has no specific
  knowledge of the IdP) to verify the assertion.  It is also possible
  to have IdPs with which the browser has a specific trust
  relationship, as described in Section 7.1.  This allows the browser
  to display a trusted element in the browser chrome indicating that a
  call is coming in from Alice.  If Alice is in Bob's address book,
  then this interface might also include her real name, a picture, etc.
  The calling site will also provide some user interface element (e.g.,
  a button) to allow Bob to answer the call, though this is most likely
  not part of the trusted UI.

  If Bob agrees, a PeerConnection is instantiated with the message from
  Alice's side.  Then, a similar process occurs as on Alice's browser:
  Bob's browser prompts him for device permission, the media streams
  are created, and a return signaling message containing media
  information, ICE candidates, and a fingerprint is sent back to Alice
  via the signaling service.  If Bob has a relationship with an IdP,
  the message will also come with an identity assertion.

  At this point, Alice and Bob each know that the other party wants to
  have a secure call with them.  Based purely on the interface provided
  by the signaling server, they know that the signaling server claims
  that the call is from Alice to Bob. This level of security is
  provided merely by having the fingerprint in the message and having
  that message received securely from the signaling server.  Because
  the far end sent an identity assertion along with their message, they
  know that this is verifiable from the IdP as well.  Note that if the
  call is federated, as shown in Figure 4, then Alice is able to verify
  Bob's identity in a way that is not mediated by either her signaling
  server or Bob's.  Rather, she verifies it directly with Bob's IdP.

  Of course, the call works perfectly well if either Alice or Bob
  doesn't have a relationship with an IdP; they just get a lower level
  of assurance.  I.e., they simply have whatever information their
  calling site claims about the caller/callee's identity.  Moreover,
  Alice might wish to make an anonymous call through an anonymous
  calling site, in which case she would of course just not provide any
  identity assertion and the calling site would mask her identity from
  Bob.

4.2.  Media Consent Verification

  As described in [RFC8826], Section 4.2, media consent verification is
  provided via ICE.  Thus, Alice and Bob perform ICE checks with each
  other.  At the completion of these checks, they are ready to send
  non-ICE data.

  At this point, Alice knows that (a) Bob (assuming he is verified via
  his IdP) or someone else who the signaling service is claiming is Bob
  is willing to exchange traffic with her and (b) either Bob is at the
  IP address which she has verified via ICE or there is an attacker who
  is on-path to that IP address detouring the traffic.  Note that it is
  not possible for an attacker who is on-path between Alice and Bob but
  not attached to the signaling service to spoof these checks because
  they do not have the ICE credentials.  Bob has the same security
  guarantees with respect to Alice.

4.3.  DTLS Handshake

  Once the requisite ICE checks have completed, Alice and Bob can set
  up a secure channel or channels.  This is performed via DTLS
  [RFC6347] and DTLS-SRTP [RFC5763] keying for SRTP [RFC3711] for the
  media channel and the Stream Control Transmission Protocol (SCTP)
  over DTLS [RFC8261] for data channels.  Specifically, Alice and Bob
  perform a DTLS handshake on every component which has been
  established by ICE.  The total number of channels depends on the
  amount of muxing; in the most likely case, we are using both RTP/RTCP
  mux and muxing multiple media streams on the same channel, in which
  case there is only one DTLS handshake.  Once the DTLS handshake has
  completed, the keys are exported [RFC5705] and used to key SRTP for
  the media channels.

  At this point, Alice and Bob know that they share a set of secure
  data and/or media channels with keys which are not known to any
  third-party attacker.  If Alice and Bob authenticated via their IdPs,
  then they also know that the signaling service is not mounting a man-
  in-the-middle attack on their traffic.  Even if they do not use an
  IdP, as long as they have minimal trust in the signaling service not
  to perform a man-in-the-middle attack, they know that their
  communications are secure against the signaling service as well
  (i.e., that the signaling service cannot mount a passive attack on
  the communications).

4.4.  Communications and Consent Freshness

  From a security perspective, everything from here on in is a little
  anticlimactic: Alice and Bob exchange data protected by the keys
  negotiated by DTLS.  Because of the security guarantees discussed in
  the previous sections, they know that the communications are
  encrypted and authenticated.

  The one remaining security property we need to establish is "consent
  freshness", i.e., allowing Alice to verify that Bob is still prepared
  to receive her communications so that Alice does not continue to send
  large traffic volumes to entities which went abruptly offline.  ICE
  specifies periodic Session Traversal Utilities for NAT (STUN)
  keepalives but only if media is not flowing.  Because the consent
  issue is more difficult here, we require WebRTC implementations to
  periodically send keepalives using the consent freshness mechanism
  specified in [RFC7675].  If a keepalive fails and no new ICE channels
  can be established, then the session is terminated.

5.  SDP Identity Attribute

  The SDP "identity" attribute is a session-level attribute that is
  used by an endpoint to convey its identity assertion to its peer.
  The identity-assertion value is encoded as base64, as described in
  Section 4 of [RFC4648].

  The procedures in this section are based on the assumption that the
  identity assertion of an endpoint is bound to the fingerprints of the
  endpoint.  This does not preclude the definition of alternative means
  of binding an assertion to the endpoint, but such means are outside
  the scope of this specification.

  The semantics of multiple "identity" attributes within an offer or
  answer are undefined.  Implementations SHOULD only include a single
  "identity" attribute in an offer or answer, and Relying Parties MAY
  elect to ignore all but the first "identity" attribute.

  Name:  identity

  Value:  identity-assertion

  Usage Level:  session

  Charset Dependent:  no

  Default Value:  N/A

  Syntax:

   identity-assertion       = identity-assertion-value
                              *(SP identity-extension)
   identity-assertion-value = base64
   identity-extension       = extension-name [ "=" extension-value ]
   extension-name           = token
   extension-value          = 1*(%x01-09 / %x0b-0c / %x0e-3a / %x3c-ff)
                              ; byte-string from [RFC4566]

   <ALPHA and DIGIT as defined in [RFC4566]>
   <base64 as defined in [RFC4566]>

  Example:

   a=identity:\
     eyJpZHAiOnsiZG9tYWluIjoiZXhhbXBsZS5vcmciLCJwcm90b2NvbCI6ImJvZ3Vz\
     In0sImFzc2VydGlvbiI6IntcImlkZW50aXR5XCI6XCJib2JAZXhhbXBsZS5vcmdc\
     IixcImNvbnRlbnRzXCI6XCJhYmNkZWZnaGlqa2xtbm9wcXJzdHV2d3l6XCIsXCJz\
     aWduYXR1cmVcIjpcIjAxMDIwMzA0MDUwNlwifSJ9

     |  Note that long lines in the example are folded to meet the
     |  column width constraints of this document; the backslash ("\")
     |  at the end of a line, the carriage return that follows, and
     |  whitespace shall be ignored.

  This specification does not define any extensions for the attribute.

  The identity-assertion value is a JSON encoded string [RFC8259].  The
  JSON object contains two keys: "assertion" and "idp".  The
  "assertion" key value contains an opaque string that is consumed by
  the IdP.  The "idp" key value contains a dictionary with one or two
  further values that identify the IdP.  See Section 7.6 for more
  details.

5.1.  Offer/Answer Considerations

  This section defines the SDP offer/answer [RFC3264] considerations
  for the SDP "identity" attribute.

  Within this section, 'initial offer' refers to the first offer in the
  SDP session that contains an SDP "identity" attribute.

5.1.1.  Generating the Initial SDP Offer

  When an offerer sends an offer, in order to provide its identity
  assertion to the peer, it includes an "identity" attribute in the
  offer.  In addition, the offerer includes one or more SDP
  "fingerprint" attributes.  The "identity" attribute MUST be bound to
  all the "fingerprint" attributes in the session description.

5.1.2.  Generating an SDP Answer

  If the answerer elects to include an "identity" attribute, it follows
  the same steps as those in Section 5.1.1.  The answerer can choose to
  include or omit an "identity" attribute independently, regardless of
  whether the offerer did so.

5.1.3.  Processing an SDP Offer or Answer

  When an endpoint receives an offer or answer that contains an
  "identity" attribute, the answerer can use the attribute information
  to contact the IdP and verify the identity of the peer.  If the
  identity requires a third-party IdP as described in Section 7.1, then
  that IdP will need to have been specifically configured.  If the
  identity verification fails, the answerer MUST discard the offer or
  answer as malformed.

5.1.4.  Modifying the Session

  When modifying a session, if the set of fingerprints is unchanged,
  then the sender MAY send the same "identity" attribute.  In this
  case, the established identity MUST be applied to existing DTLS
  connections as well as new connections established using one of those
  fingerprints.  Note that [RFC8829], Section 5.2.1 requires that each
  media section use the same set of fingerprints.  If a new "identity"
  attribute is received, then the receiver MUST apply that identity to
  all existing connections.

  If the set of fingerprints changes, then the sender MUST either send
  a new "identity" attribute or none at all.  Because a change in
  fingerprints also causes a new DTLS connection to be established, the
  receiver MUST discard all previously established identities.

6.  Detailed Technical Description

6.1.  Origin and Web Security Issues

  The basic unit of permissions for WebRTC is the origin [RFC6454].
  Because the security of the origin depends on being able to
  authenticate content from that origin, the origin can only be
  securely established if data is transferred over HTTPS [RFC2818].
  Thus, clients MUST treat HTTP and HTTPS origins as different
  permissions domains.  Note: This follows directly from the origin
  security model and is stated here merely for clarity.

  Many Web browsers currently forbid by default any active mixed
  content on HTTPS pages.  That is, when JavaScript is loaded from an
  HTTP origin onto an HTTPS page, an error is displayed and the HTTP
  content is not executed unless the user overrides the error.  Any
  browser which enforces such a policy will also not permit access to
  WebRTC functionality from mixed content pages (because they never
  display mixed content).  Browsers which allow active mixed content
  MUST nevertheless disable WebRTC functionality in mixed content
  settings.

  Note that it is possible for a page which was not mixed content to
  become mixed content during the duration of the call.  The major risk
  here is that the newly arrived insecure JS might redirect media to a
  location controlled by the attacker.  Implementations MUST either
  choose to terminate the call or display a warning at that point.

  Also note that the security architecture depends on the keying
  material not being available to move between origins.  However, it is
  assumed that the identity assertion can be passed to anyone that the
  page cares to.

6.2.  Device Permissions Model

  Implementations MUST obtain explicit user consent prior to providing
  access to the camera and/or microphone.  Implementations MUST at
  minimum support the following two permissions models for HTTPS
  origins.

  *  Requests for one-time camera/microphone access.

  *  Requests for permanent access.

  Because HTTP origins cannot be securely established against network
  attackers, implementations MUST refuse all permissions grants for
  HTTP origins.

  In addition, they SHOULD support requests for access that promise
  that media from this grant will be sent to a single communicating
  peer (obviously there could be other requests for other peers), e.g.,
  "Call [email protected]".  The semantics of this request
  are that the media stream from the camera and microphone will only be
  routed through a connection which has been cryptographically verified
  (through the IdP mechanism or an X.509 certificate in the DTLS-SRTP
  handshake) as being associated with the stated identity.  Note that
  it is unlikely that browsers would have X.509 certificates, but
  servers might.  Browsers servicing such requests SHOULD clearly
  indicate that identity to the user when asking for permission.  The
  idea behind this type of permissions is that a user might have a
  fairly narrow list of peers they are willing to communicate with,
  e.g., "my mother" rather than "anyone on Facebook".  Narrow
  permissions grants allow the browser to do that enforcement.

  API Requirement:  The API MUST provide a mechanism for the requesting
     JS to relinquish the ability to see or modify the media (e.g., via
     MediaStream.record()).  Combined with secure authentication of the
     communicating peer, this allows a user to be sure that the calling
     site is not accessing or modifying their conversion.

  UI Requirement:  The UI MUST clearly indicate when the user's camera
     and microphone are in use.  This indication MUST NOT be
     suppressible by the JS and MUST clearly indicate how to terminate
     device access, and provide a UI means to immediately stop camera/
     microphone input without the JS being able to prevent it.

  UI Requirement:  If the UI indication of camera/microphone use is
     displayed in the browser such that minimizing the browser window
     would hide the indication, or the JS creating an overlapping
     window would hide the indication, then the browser SHOULD stop
     camera and microphone input when the indication is hidden.  (Note:
     This may not be necessary in systems that are non-windows-based
     but that have good notifications support, such as phones.)

  *  Browsers MUST NOT permit permanent screen or application sharing
     permissions to be installed as a response to a JS request for
     permissions.  Instead, they must require some other user action
     such as a permissions setting or an application install experience
     to grant permission to a site.

  *  Browsers MUST provide a separate dialog request for screen/
     application sharing permissions even if the media request is made
     at the same time as the request for camera and microphone
     permissions.

  *  The browser MUST indicate any windows which are currently being
     shared in some unambiguous way.  Windows which are not visible
     MUST NOT be shared even if the application is being shared.  If
     the screen is being shared, then that MUST be indicated.

  Browsers MAY permit the formation of data channels without any direct
  user approval.  Because sites can always tunnel data through the
  server, further restrictions on the data channel do not provide any
  additional security.  (See Section 6.3 for a related issue.)

  Implementations which support some form of direct user authentication
  SHOULD also provide a policy by which a user can authorize calls only
  to specific communicating peers.  Specifically, the implementation
  SHOULD provide the following interfaces/controls:

  *  Allow future calls to this verified user.

  *  Allow future calls to any verified user who is in my system
     address book (this only works with address book integration, of
     course).

  Implementations SHOULD also provide a different user interface
  indication when calls are in progress to users whose identities are
  directly verifiable.  Section 6.5 provides more on this.

6.3.  Communications Consent

  Browser client implementations of WebRTC MUST implement ICE.  Server
  gateway implementations which operate only at public IP addresses
  MUST implement either full ICE or ICE-Lite [RFC8445].

  Browser implementations MUST verify reachability via ICE prior to
  sending any non-ICE packets to a given destination.  Implementations
  MUST NOT provide the ICE transaction ID to JavaScript during the
  lifetime of the transaction (i.e., during the period when the ICE
  stack would accept a new response for that transaction).  The JS MUST
  NOT be permitted to control the local ufrag and password, though it
  of course knows it.

  While continuing consent is required, the ICE [RFC8445], Section 11
  keepalives use STUN Binding Indications, which are one-way and
  therefore not sufficient.  The current WG consensus is to use ICE
  Binding Requests for continuing consent freshness.  ICE already
  requires that implementations respond to such requests, so this
  approach is maximally compatible.  A separate document will profile
  the ICE timers to be used; see [RFC7675].

6.4.  IP Location Privacy

  A side effect of the default ICE behavior is that the peer learns
  one's IP address, which leaks large amounts of location information.
  This has negative privacy consequences in some circumstances.  The
  API requirements in this section are intended to mitigate this issue.
  Note that these requirements are not intended to protect the user's
  IP address from a malicious site.  In general, the site will learn at
  least a user's server-reflexive address from any HTTP transaction.
  Rather, these requirements are intended to allow a site to cooperate
  with the user to hide the user's IP address from the other side of
  the call.  Hiding the user's IP address from the server requires some
  sort of explicit privacy-preserving mechanism on the client (e.g.,
  Tor Browser <https://www.torproject.org/projects/torbrowser.html.en>)
  and is out of scope for this specification.

  API Requirement:  The API MUST provide a mechanism to allow the JS to
     suppress ICE negotiation (though perhaps to allow candidate
     gathering) until the user has decided to answer the call.  (Note:
     Determining when the call has been answered is a question for the
     JS.)  This enables a user to prevent a peer from learning their IP
     address if they elect not to answer a call and also from learning
     whether the user is online.

  API Requirement:  The API MUST provide a mechanism for the calling
     application JS to indicate that only TURN candidates are to be
     used.  This prevents the peer from learning one's IP address at
     all.  This mechanism MUST also permit suppression of the related
     address field, since that leaks local addresses.

  API Requirement:  The API MUST provide a mechanism for the calling
     application to reconfigure an existing call to add non-TURN
     candidates.  Taken together, this and the previous requirement
     allow ICE negotiation to start immediately on incoming call
     notification, thus reducing post-dial delay, but also to avoid
     disclosing the user's IP address until they have decided to
     answer.  They also allow users to completely hide their IP address
     for the duration of the call.  Finally, they allow a mechanism for
     the user to optimize performance by reconfiguring to allow non-
     TURN candidates during an active call if the user decides they no
     longer need to hide their IP address.

  Note that some enterprises may operate proxies and/or NATs designed
  to hide internal IP addresses from the outside world.  WebRTC
  provides no explicit mechanism to allow this function.  Either such
  enterprises need to proxy the HTTP/HTTPS and modify the SDP and/or
  the JS, or there needs to be browser support to set the "TURN-only"
  policy regardless of the site's preferences.

  Note: These requirements are intended to allow sites to conceal the
  user's IP address from the peer.  For guidance on concealing the
  user's IP address from the calling site see [RFC8828].

6.5.  Communications Security

  Implementations MUST support SRTP [RFC3711].  Implementations MUST
  support DTLS [RFC6347] and DTLS-SRTP [RFC5763] [RFC5764] for SRTP
  keying.  Implementations MUST support SCTP over DTLS [RFC8261].

  All media channels MUST be secured via SRTP and the Secure Real-time
  Transport Control Protocol (SRTCP).  Media traffic MUST NOT be sent
  over plain (unencrypted) RTP or RTCP; that is, implementations MUST
  NOT negotiate cipher suites with NULL encryption modes.  DTLS-SRTP
  MUST be offered for every media channel.  WebRTC implementations MUST
  NOT offer SDP security descriptions [RFC4568] or select it if
  offered.  An SRTP Master Key Identifier (MKI) MUST NOT be used.

  All data channels MUST be secured via DTLS.

  All implementations MUST support DTLS 1.2 with the
  TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256 cipher suite and the P-256
  curve [FIPS186].  Earlier drafts of this specification required DTLS
  1.0 with the cipher suite TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA, and
  at the time of this writing some implementations do not support DTLS
  1.2; endpoints which support only DTLS 1.2 might encounter
  interoperability issues.  The DTLS-SRTP protection profile
  SRTP_AES128_CM_HMAC_SHA1_80 MUST be supported for SRTP.
  Implementations MUST favor cipher suites which support Forward
  Secrecy (FS) over non-FS cipher suites and SHOULD favor Authenticated
  Encryption with Associated Data (AEAD) over non-AEAD cipher suites.
  Note: the IETF is in the process of standardizing DTLS 1.3
  [TLS-DTLS13].

  Implementations MUST NOT implement DTLS renegotiation and MUST reject
  it with a "no_renegotiation" alert if offered.

  Endpoints MUST NOT implement TLS False Start [RFC7918].

  API Requirement:  The API MUST generate a new authentication key pair
     for every new call by default.  This is intended to allow for
     unlinkability.

  API Requirement:  The API MUST provide a means to reuse a key pair
     for calls.  This can be used to enable key continuity-based
     authentication, and could be used to amortize key generation
     costs.

  API Requirement:  Unless the user specifically configures an external
     key pair, different key pairs MUST be used for each origin.  (This
     avoids creating a super-cookie.)

  API Requirement:  When DTLS-SRTP is used, the API MUST NOT permit the
     JS to obtain the negotiated keying material.  This requirement
     preserves the end-to-end security of the media.

  UI Requirements:  A user-oriented client MUST provide an "inspector"
     interface which allows the user to determine the "security
     characteristics" of the media.

     The following properties SHOULD be displayed "up-front" in the
     browser chrome, i.e., without requiring the user to ask for them:


     *  A client MUST provide a user interface through which a user may
        determine the "security characteristics" for currently
        displayed audio and video stream(s).

     *  A client MUST provide a user interface through which a user may
        determine the "security characteristics" for transmissions of
        their microphone audio and camera video.

     *  If the far endpoint was directly verified, either via a third-
        party verifiable X.509 certificate or via a Web IdP mechanism
        (see Section 7), the "security characteristics" MUST include
        the verified information.  X.509 identities and Web IdP
        identities have similar semantics and should be displayed in a
        similar way.

     The following properties are more likely to require some "drill-
     down" from the user:


     *  The "security characteristics" MUST indicate the cryptographic
        algorithms in use (for example, "AES-CBC").

     *  The "security characteristics" MUST indicate whether FS is
        provided.

     *  The "security characteristics" MUST include some mechanism to
        allow an out-of-band verification of the peer, such as a
        certificate fingerprint or a Short Authentication String (SAS).
        These are compared by the peers to authenticate one another.

7.  Web-Based Peer Authentication

  NOTE: The mechanism described in this section was designed relatively
  early in the RTCWEB process.  In retrospect, the WG was too
  optimistic about the enthusiasm for this kind of mechanism.  At the
  time of publication, it has not been widely adopted or implemented.
  It appears in this document as a description of the state of the art
  as of this writing.

  In a number of cases, it is desirable for the endpoint (i.e., the
  browser) to be able to directly identify the endpoint on the other
  side without trusting the signaling service to which they are
  connected.  For instance, users may be making a call via a federated
  system where they wish to get direct authentication of the other
  side.  Alternately, they may be making a call on a site which they
  minimally trust (such as a poker site) but to someone who has an
  identity on a site they do trust (such as a social network).

  Recently, a number of Web-based identity technologies (OAuth,
  Facebook Connect, etc.) have been developed.  While the details vary,
  what these technologies share is that they have a Web-based (i.e.,
  HTTP/HTTPS) IdP which attests to Alice's identity.  For instance, if
  Alice has an account at example.org, Alice could use the example.org
  IdP to prove to others that Alice is [email protected].  The
  development of these technologies allows us to separate calling from
  identity provision: Alice could call you on a poker site but identify
  herself as [email protected].

  Whatever the underlying technology, the general principle is that the
  party which is being authenticated is NOT the signaling site but
  rather the user (and their browser).  Similarly, the Relying Party is
  the browser and not the signaling site.  Thus, the browser MUST
  generate the input to the IdP assertion process and display the
  results of the verification process to the user in a way which cannot
  be imitated by the calling site.

  The mechanisms defined in this document do not require the browser to
  implement any particular identity protocol or to support any
  particular IdP.  Instead, this document provides a generic interface
  which any IdP can implement.  Thus, new IdPs and protocols can be
  introduced without change to either the browser or the calling
  service.  This avoids the need to make a commitment to any particular
  identity protocol, although browsers may opt to directly implement
  some identity protocols in order to provide superior performance or
  UI properties.

7.1.  Trust Relationships: IdPs, APs, and RPs

  Any federated identity protocol has three major participants:

  Authenticating Party (AP):  The entity which is trying to establish
     its identity.

  Identity Provider (IdP):  The entity which is vouching for the AP's
     identity.

  Relying Party (RP):  The entity which is trying to verify the AP's
     identity.

  The AP and the IdP have an account relationship of some kind: the AP
  registers with the IdP and is able to subsequently authenticate
  directly to the IdP (e.g., with a password).  This means that the
  browser must somehow know which IdP(s) the user has an account
  relationship with.  This can either be something that the user
  configures into the browser or that is configured at the calling site
  and then provided to the PeerConnection by the Web application at the
  calling site.  The use case for having this information configured
  into the browser is that the user may "log into" the browser to bind
  it to some identity.  This is becoming common in new browsers.
  However, it should also be possible for the IdP information to simply
  be provided by the calling application.

  At a high level, there are two kinds of IdPs:

  Authoritative:  IdPs which have verifiable control of some section of
     the identity space.  For instance, in the realm of email, the
     operator of "example.com" has complete control of the namespace
     ending in "@example.com".  Thus, "[email protected]" is whoever
     the operator says it is.  Examples of systems with authoritative
     IdPs include DNSSEC, an identity system for SIP (see [RFC8224]),
     and Facebook Connect (Facebook identities only make sense within
     the context of the Facebook system).

  Third-Party:  IdPs which don't have control of their section of the
     identity space but instead verify users' identities via some
     unspecified mechanism and then attest to it.  Because the IdP
     doesn't actually control the namespace, RPs need to trust that the
     IdP is correctly verifying AP identities, and there can
     potentially be multiple IdPs attesting to the same section of the
     identity space.  Probably the best-known example of a third-party
     IdP is SSL/TLS certificates, where there are a large number of
     certificate authorities (CAs) all of whom can attest to any domain
     name.

  If an AP is authenticating via an authoritative IdP, then the RP does
  not need to explicitly configure trust in the IdP at all.  The
  identity mechanism can directly verify that the IdP indeed made the
  relevant identity assertion (a function provided by the mechanisms in
  this document), and any assertion it makes about an identity for
  which it is authoritative is directly verifiable.  Note that this
  does not mean that the IdP might not lie, but that is a
  trustworthiness judgement that the user can make at the time they
  look at the identity.

  By contrast, if an AP is authenticating via a third-party IdP, the RP
  needs to explicitly trust that IdP (hence the need for an explicit
  trust anchor list in PKI-based SSL/TLS clients).  The list of
  trustable IdPs needs to be configured directly into the browser,
  either by the user or potentially by the browser manufacturer.  This
  is a significant advantage of authoritative IdPs and implies that if
  third-party IdPs are to be supported, the potential number needs to
  be fairly small.

7.2.  Overview of Operation

  In order to provide security without trusting the calling site, the
  PeerConnection component of the browser must interact directly with
  the IdP.  The details of the mechanism are described in the W3C API
  specification, but the general idea is that the PeerConnection
  component downloads JS from a specific location on the IdP dictated
  by the IdP domain name.  That JS (the "IdP proxy") runs in an
  isolated security context within the browser, and the PeerConnection
  talks to it via a secure message passing channel.

  Note that there are two logically separate functions here:

  *  Identity assertion generation.

  *  Identity assertion verification.

  The same IdP JS "endpoint" is used for both functions, but of course
  a given IdP might behave differently and load new JS to perform one
  function or the other.

       +--------------------------------------+
       | Browser                              |
       |                                      |
       | +----------------------------------+ |
       | | https://calling-site.example.com | |
       | |                                  | |
       | |        Calling JS Code           | |
       | |               ^                  | |
       | +---------------|------------------+ |
       |                 | API Calls          |
       |                 v                    |
       |          PeerConnection              |
       |                 ^                    |
       |                 | API Calls          |
       |     +-----------|-------------+      |   +---------------+
       |     |           v             |      |   |               |
       |     |       IdP Proxy         |<-------->|   Identity    |
       |     |                         |      |   |   Provider    |
       |     | https://idp.example.org |      |   |               |
       |     +-------------------------+      |   +---------------+
       |                                      |
       +--------------------------------------+

  When the PeerConnection object wants to interact with the IdP, the
  sequence of events is as follows:

  1.  The browser (the PeerConnection component) instantiates an IdP
      proxy.  This allows the IdP to load whatever JS is necessary into
      the proxy.  The resulting code runs in the IdP's security
      context.

  2.  The IdP registers an object with the browser that conforms to the
      API defined in [webrtc-api].

  3.  The browser invokes methods on the object registered by the IdP
      proxy to create or verify identity assertions.

  This approach allows us to decouple the browser from any particular
  IdP; the browser need only know how to load the IdP's JavaScript --
  the location of which is determined based on the IdP's identity --
  and to call the generic API for requesting and verifying identity
  assertions.  The IdP provides whatever logic is necessary to bridge
  the generic protocol to the IdP's specific requirements.  Thus, a
  single browser can support any number of identity protocols,
  including being forward compatible with IdPs which did not exist at
  the time the browser was written.

7.3.  Items for Standardization

  There are two parts to this work:

  *  The precise information from the signaling message that must be
     cryptographically bound to the user's identity and a mechanism for
     carrying assertions in JavaScript Session Establishment Protocol
     (JSEP) messages.  This is specified in Section 7.4.

  *  The interface to the IdP, which is defined in the companion W3C
     WebRTC API specification [webrtc-api].

  The WebRTC API specification also defines JavaScript interfaces that
  the calling application can use to specify which IdP to use.  That
  API also provides access to the assertion-generation capability and
  the status of the validation process.

7.4.  Binding Identity Assertions to JSEP Offer/Answer Transactions

  An identity assertion binds the user's identity (as asserted by the
  IdP) to the SDP offer/answer exchange and specifically to the media.
  In order to achieve this, the PeerConnection must provide the DTLS-
  SRTP fingerprint to be bound to the identity.  This is provided as a
  JavaScript object (also known as a dictionary or hash) with a single
  "fingerprint" key, as shown below:

  {
    "fingerprint":
      [
        { "algorithm": "sha-256",
          "digest": "4A:AD:B9:B1:3F:...:E5:7C:AB" },
        { "algorithm": "sha-1",
          "digest": "74:E9:76:C8:19:...:F4:45:6B" }
      ]
  }

  The "fingerprint" value is an array of objects.  Each object in the
  array contains "algorithm" and "digest" values, which correspond
  directly to the algorithm and digest values in the "fingerprint"
  attribute of the SDP [RFC8122].

  This object is encoded in a JSON [RFC8259] string for passing to the
  IdP.  The identity assertion returned by the IdP, which is encoded in
  the "identity" attribute, is a JSON object that is encoded as
  described in Section 7.4.1.

  This structure does not need to be interpreted by the IdP or the IdP
  proxy.  It is consumed solely by the RP's browser.  The IdP merely
  treats it as an opaque value to be attested to.  Thus, new parameters
  can be added to the assertion without modifying the IdP.

7.4.1.  Carrying Identity Assertions

  Once an IdP has generated an assertion (see Section 7.6), it is
  attached to the SDP offer/answer message.  This is done by adding a
  new "identity" attribute to the SDP.  The sole contents of this value
  is the identity assertion.  The identity assertion produced by the
  IdP is encoded into a UTF-8 JSON text, then base64-encoded [RFC4648]
  to produce this string.  For example:

  v=0
  o=- 1181923068 1181923196 IN IP4 ua1.example.com
  s=example1
  c=IN IP4 ua1.example.com
  a=fingerprint:sha-1 \
    4A:AD:B9:B1:3F:82:18:3B:54:02:12:DF:3E:5D:49:6B:19:E5:7C:AB
  a=identity:\
    eyJpZHAiOnsiZG9tYWluIjoiZXhhbXBsZS5vcmciLCJwcm90b2NvbCI6ImJvZ3Vz\
    In0sImFzc2VydGlvbiI6IntcImlkZW50aXR5XCI6XCJib2JAZXhhbXBsZS5vcmdc\
    IixcImNvbnRlbnRzXCI6XCJhYmNkZWZnaGlqa2xtbm9wcXJzdHV2d3l6XCIsXCJz\
    aWduYXR1cmVcIjpcIjAxMDIwMzA0MDUwNlwifSJ9
  a=...
  t=0 0
  m=audio 6056 RTP/SAVP 0
  a=sendrecv
  ...

     |  Note that long lines in the example are folded to meet the
     |  column width constraints of this document; the backslash ("\")
     |  at the end of a line, the carriage return that follows, and
     |  whitespace shall be ignored.

  The "identity" attribute attests to all "fingerprint" attributes in
  the session description.  It is therefore a session-level attribute.

  Multiple "fingerprint" values can be used to offer alternative
  certificates for a peer.  The "identity" attribute MUST include all
  "fingerprint" values that are included in "fingerprint" attributes of
  the session description.

  The RP browser MUST verify that the in-use certificate for a DTLS
  connection is in the set of fingerprints returned from the IdP when
  verifying an assertion.

7.5.  Determining the IdP URI

  In order to ensure that the IdP is under control of the domain owner
  rather than someone who merely has an account on the domain owner's
  server (e.g., in shared hosting scenarios), the IdP JavaScript is
  hosted at a deterministic location based on the IdP's domain name.
  Each IdP proxy instance is associated with two values:

  authority:  The authority [RFC3986] at which the IdP's service is
     hosted.

  protocol:  The specific IdP protocol which the IdP is using.  This is
     a completely opaque IdP-specific string, but allows an IdP to
     implement two protocols in parallel.  This value may be the empty
     string.  If no value for protocol is provided, a value of
     "default" is used.

  Each IdP MUST serve its initial entry page (i.e., the one loaded by
  the IdP proxy) from a well-known URI [RFC8615].  The well-known URI
  for an IdP proxy is formed from the following URI components:

  1.  The scheme, "https:".  An IdP MUST be loaded using HTTPS
      [RFC2818].

  2.  The authority [RFC3986].  As noted above, the authority MAY
      contain a non-default port number or userinfo sub-component.
      Both are removed when determining if an asserted identity matches
      the name of the IdP.

  3.  The path, starting with "/.well-known/idp-proxy/" and appended
      with the IdP protocol.  Note that the separator characters '/'
      (%2F) and '\' (%5C) MUST NOT be permitted in the protocol field,
      lest an attacker be able to direct requests outside of the
      controlled "/.well-known/" prefix.  Query and fragment values MAY
      be used by including '?' or '#' characters.

  For example, for the IdP "identity.example.com" and the protocol
  "example", the URL would be:

  https://identity.example.com/.well-known/idp-proxy/example

  The IdP MAY redirect requests to this URL, but they MUST retain the
  "https:" scheme.  This changes the effective origin of the IdP, but
  not the domain of the identities that the IdP is permitted to assert
  and validate.  I.e., the IdP is still regarded as authoritative for
  the original domain.

7.5.1.  Authenticating Party

  How an AP determines the appropriate IdP domain is out of scope of
  this specification.  In general, however, the AP has some actual
  account relationship with the IdP, as this identity is what the IdP
  is attesting to.  Thus, the AP somehow supplies the IdP information
  to the browser.  Some potential mechanisms include:

  *  Provided by the user directly.

  *  Selected from some set of IdPs known to the calling site (e.g., a
     button that shows "Authenticate via Facebook Connect").

7.5.2.  Relying Party

  Unlike the AP, the RP need not have any particular relationship with
  the IdP.  Rather, it needs to be able to process whatever assertion
  is provided by the AP.  As the assertion contains the IdP's identity
  in the "idp" field of the JSON-encoded object (see Section 7.6), the
  URI can be constructed directly from the assertion, and thus the RP
  can directly verify the technical validity of the assertion with no
  user interaction.  Authoritative assertions need only be verifiable.
  Third-party assertions also MUST be verified against local policy, as
  described in Section 8.1.

7.6.  Requesting Assertions

  The input to the identity assertion generation process is the JSON-
  encoded object described in Section 7.4 that contains the set of
  certificate fingerprints the browser intends to use.  This string is
  treated as opaque from the perspective of the IdP.

  The browser also identifies the origin that the PeerConnection is run
  in, which allows the IdP to make decisions based on who is requesting
  the assertion.

  An application can optionally provide a user identifier hint when
  specifying an IdP.  This value is a hint that the IdP can use to
  select amongst multiple identities, or to avoid providing assertions
  for unwanted identities.  The "username" is a string that has no
  meaning to any entity other than the IdP; it can contain any data the
  IdP needs in order to correctly generate an assertion.

  An identity assertion that is successfully provided by the IdP
  consists of the following information:

  idp:  The domain name of an IdP and the protocol string.  This MAY
     identify a different IdP or protocol from the one that generated
     the assertion.

  assertion:  An opaque value containing the assertion itself.  This is
     only interpretable by the identified IdP or the IdP code running
     in the client.

  Figure 5 shows an example assertion formatted as JSON.  In this case,
  the message has presumably been digitally signed/MACed in some way
  that the IdP can later verify it, but this is an implementation
  detail and out of scope of this document.

  {
    "idp":{
      "domain": "example.org",
      "protocol": "bogus"
    },
    "assertion": "{\"identity\":\"[email protected]\",
                   \"contents\":\"abcdefghijklmnopqrstuvwyz\",
                   \"signature\":\"010203040506\"}"
  }

                       Figure 5: Example Assertion

  For use in signaling, the assertion is serialized into JSON,
  base64-encoded [RFC4648], and used as the value of the "identity"
  attribute.  IdPs SHOULD ensure that any assertions they generate
  cannot be interpreted in a different context.  E.g., they should use
  a distinct format or have separate cryptographic keys for assertion
  generation and other purposes.  Line breaks are inserted solely for
  readability.

7.7.  Managing User Login

  In order to generate an identity assertion, the IdP needs proof of
  the user's identity.  It is common practice to authenticate users
  (using passwords or multi-factor authentication), then use cookies
  [RFC6265] or HTTP authentication [RFC7617] for subsequent exchanges.

  The IdP proxy is able to access cookies, HTTP authentication data, or
  other persistent session data because it operates in the security
  context of the IdP origin.  Therefore, if a user is logged in, the
  IdP could have all the information needed to generate an assertion.

  An IdP proxy is unable to generate an assertion if the user is not
  logged in, or the IdP wants to interact with the user to acquire more
  information before generating the assertion.  If the IdP wants to
  interact with the user before generating an assertion, the IdP proxy
  can fail to generate an assertion and instead indicate a URL where
  login should proceed.

  The application can then load the provided URL to enable the user to
  enter credentials.  The communication between the application and the
  IdP is described in [webrtc-api].

8.  Verifying Assertions

  The input to identity validation is the assertion string taken from a
  decoded "identity" attribute.

  The IdP proxy verifies the assertion.  Depending on the identity
  protocol, the proxy might contact the IdP server or other servers.
  For instance, an OAuth-based protocol will likely require using the
  IdP as an oracle, whereas with a signature-based scheme it might be
  able to verify the assertion without contacting the IdP, provided
  that it has cached the relevant public key.

  Regardless of the mechanism, if verification succeeds, a successful
  response from the IdP proxy consists of the following information:

  identity:  The identity of the AP from the IdP's perspective.
     Details of this are provided in Section 8.1.

  contents:  The original unmodified string provided by the AP as input
     to the assertion generation process.

  Figure 6 shows an example response, which is JSON-formatted.

  {
    "identity": "[email protected]",
    "contents": "{\"fingerprint\":[ ... ]}"
  }

                  Figure 6: Example Verification Result

8.1.  Identity Formats

  The identity provided from the IdP to the RP browser MUST consist of
  a string representing the user's identity.  This string is in the
  form "<user>@<domain>", where "user" consists of any character, and
  domain is an internationalized domain name [RFC5890] encoded as a
  sequence of U-labels.

  The PeerConnection API MUST check this string as follows:

  1.  If the "domain" portion of the string is equal to the domain name
      of the IdP proxy, then the assertion is valid, as the IdP is
      authoritative for this domain.  Comparison of domain names is
      done using the label equivalence rule defined in Section 2.3.2.4
      of [RFC5890].

  2.  If the "domain" portion of the string is not equal to the domain
      name of the IdP proxy, then the PeerConnection object MUST reject
      the assertion unless both:

      1.  the IdP domain is trusted as an acceptable third-party IdP;
          and

      2.  local policy is configured to trust this IdP domain for the
          domain portion of the identity string.

  Any '@' or '%' characters in the "user" portion of the identity MUST
  be escaped according to the "percent-encoding" rules defined in
  Section 2.1 of [RFC3986].  Characters other than '@' and '%' MUST NOT
  be percent-encoded.  For example, with a "user" of "user@133" and a
  "domain" of "identity.example.com", the resulting string will be
  encoded as "user%[email protected]".

  Implementations are cautioned to take care when displaying user
  identities containing escaped '@' characters.  If such characters are
  unescaped prior to display, implementations MUST distinguish between
  the domain of the IdP proxy and any domain that might be implied by
  the portion of the "<user>" portion that appears after the escaped
  "@" sign.

9.  Security Considerations

  Much of the security analysis of RTCWEB is contained in [RFC8826] or
  in the discussion of the particular issues above.  In order to avoid
  repetition, this section focuses on (a) residual threats that are not
  addressed by this document and (b) threats produced by failure/
  misbehavior of one of the components in the system.

9.1.  Communications Security

  If HTTPS is not used to secure communications to the signaling
  server, and the identity mechanism used in Section 7 is not used,
  then any on-path attacker can replace the DTLS-SRTP fingerprints in
  the handshake and thus substitute its own identity for that of either
  endpoint.

  Even if HTTPS is used, the signaling server can potentially mount a
  man-in-the-middle attack unless implementations have some mechanism
  for independently verifying keys.  The UI requirements in Section 6.5
  are designed to provide such a mechanism for motivated/security
  conscious users, but are not suitable for general use.  The identity
  service mechanisms in Section 7 are more suitable for general use.
  Note, however, that a malicious signaling service can strip off any
  such identity assertions, though it cannot forge new ones.  Note that
  all of the third-party security mechanisms available (whether X.509
  certificates or a third-party IdP) rely on the security of the third
  party -- this is of course also true of the user's connection to the
  Web site itself.  Users who wish to assure themselves of security
  against a malicious IdP can only do so by verifying peer credentials
  directly, e.g., by checking the peer's fingerprint against a value
  delivered out of band.

  In order to protect against malicious content JavaScript, that
  JavaScript MUST NOT be allowed to have direct access to -- or perform
  computations with -- DTLS keys.  For instance, if content JS were
  able to compute digital signatures, then it would be possible for
  content JS to get an identity assertion for a browser's generated key
  and then use that assertion plus a signature by the key to
  authenticate a call protected under an ephemeral Diffie-Hellman (DH)
  key controlled by the content JS, thus violating the security
  guarantees otherwise provided by the IdP mechanism.  Note that it is
  not sufficient merely to deny the content JS direct access to the
  keys, as some have suggested doing with the WebCrypto API
  [webcrypto].  The JS must also not be allowed to perform operations
  that would be valid for a DTLS endpoint.  By far the safest approach
  is simply to deny the ability to perform any operations that depend
  on secret information associated with the key.  Operations that
  depend on public information, such as exporting the public key, are
  of course safe.

9.2.  Privacy

  The requirements in this document are intended to allow:

  *  Users to participate in calls without revealing their location.

  *  Potential callees to avoid revealing their location and even
     presence status prior to agreeing to answer a call.

  However, these privacy protections come at a performance cost in
  terms of using TURN relays and, in the latter case, delaying ICE.
  Sites SHOULD make users aware of these tradeoffs.

  Note that the protections provided here assume a non-malicious
  calling service.  As the calling service always knows the user's
  status and (absent the use of a technology like Tor) their IP
  address, they can violate the user's privacy at will.  Users who wish
  privacy against the calling sites they are using must use separate
  privacy-enhancing technologies such as Tor. Combined WebRTC/Tor
  implementations SHOULD arrange to route the media as well as the
  signaling through Tor. Currently this will produce very suboptimal
  performance.

  Additionally, any identifier which persists across multiple calls is
  potentially a problem for privacy, especially for anonymous calling
  services.  Such services SHOULD instruct the browser to use separate
  DTLS keys for each call and also to use TURN throughout the call.
  Otherwise, the other side will learn linkable information that would
  allow them to correlate the browser across multiple calls.
  Additionally, browsers SHOULD implement the privacy-preserving CNAME
  generation mode of [RFC7022].

9.3.  Denial of Service

  The consent mechanisms described in this document are intended to
  mitigate denial-of-service (DoS) attacks in which an attacker uses
  clients to send large amounts of traffic to a victim without the
  consent of the victim.  While these mechanisms are sufficient to
  protect victims who have not implemented WebRTC at all, WebRTC
  implementations need to be more careful.

  Consider the case of a call center which accepts calls via WebRTC.
  An attacker proxies the call center's front-end and arranges for
  multiple clients to initiate calls to the call center.  Note that
  this requires user consent in many cases, but because the data
  channel does not need consent, they can use that directly.  Since ICE
  will complete, browsers can then be induced to send large amounts of
  data to the victim call center if it supports the data channel at
  all.  Preventing this attack requires that automated WebRTC
  implementations implement sensible flow control and have the ability
  to triage out (i.e., stop responding to ICE probes on) calls which
  are behaving badly, and especially to be prepared to remotely
  throttle the data channel in the absence of plausible audio and video
  (which the attacker cannot control).

  Another related attack is for the signaling service to swap the ICE
  candidates for the audio and video streams, thus forcing a browser to
  send video to the sink that the other victim expects will contain
  audio (perhaps it is only expecting audio!), potentially causing
  overload.  Muxing multiple media flows over a single transport makes
  it harder to individually suppress a single flow by denying ICE
  keepalives.  Either media-level (RTCP) mechanisms must be used or the
  implementation must deny responses entirely, thus terminating the
  call.

  Yet another attack, suggested by Magnus Westerlund, is for the
  attacker to cross-connect offers and answers as follows.  It induces
  the victim to make a call and then uses its control of other users'
  browsers to get them to attempt a call to someone.  It then
  translates their offers into apparent answers to the victim, which
  looks like large-scale parallel forking.  The victim still responds
  to ICE responses, and now the browsers all try to send media to the
  victim.  Implementations can defend themselves from this attack by
  only responding to ICE Binding Requests for a limited number of
  remote ufrags (this is the reason for the requirement that the JS not
  be able to control the ufrag and password).  [RFC8834], Section 13
  documents a number of potential RTCP-based DoS attacks and
  countermeasures.

  Note that attacks based on confusing one end or the other about
  consent are possible even in the face of the third-party identity
  mechanism as long as major parts of the signaling messages are not
  signed.  On the other hand, signing the entire message severely
  restricts the capabilities of the calling application, so there are
  difficult tradeoffs here.

9.4.  IdP Authentication Mechanism

  This mechanism relies for its security on the IdP and on the
  PeerConnection correctly enforcing the security invariants described
  above.  At a high level, the IdP is attesting that the user
  identified in the assertion wishes to be associated with the
  assertion.  Thus, it must not be possible for arbitrary third parties
  to get assertions tied to a user or to produce assertions that RPs
  will accept.

9.4.1.  PeerConnection Origin Check

  Fundamentally, the IdP proxy is just a piece of HTML and JS loaded by
  the browser, so nothing stops a Web attacker from creating their own
  IFRAME, loading the IdP proxy HTML/JS, and requesting a signature
  over their own keys rather than those generated in the browser.
  However, that proxy would be in the attacker's origin, not the IdP's
  origin.  Only the browser itself can instantiate a context that
  (a) is in the IdP's origin and (b) exposes the correct API surface.
  Thus, the IdP proxy on the sender's side MUST ensure that it is
  running in the IdP's origin prior to issuing assertions.

  Note that this check only asserts that the browser (or some other
  entity with access to the user's authentication data) attests to the
  request and hence to the fingerprint.  It does not demonstrate that
  the browser has access to the associated private key, and therefore
  an attacker can attach their own identity to another party's keying
  material, thus making a call which comes from Alice appear to come
  from the attacker.  See [RFC8844] for defenses against this form of
  attack.

9.4.2.  IdP Well-Known URI

  As described in Section 7.5, the IdP proxy HTML/JS landing page is
  located at a well-known URI based on the IdP's domain name.  This
  requirement prevents an attacker who can write some resources at the
  IdP (e.g., on one's Facebook wall) from being able to impersonate the
  IdP.

9.4.3.  Privacy of IdP-Generated Identities and the Hosting Site

  Depending on the structure of the IdP's assertions, the calling site
  may learn the user's identity from the perspective of the IdP.  In
  many cases, this is not an issue because the user is authenticating
  to the site via the IdP in any case -- for instance, when the user
  has logged in with Facebook Connect and is then authenticating their
  call with a Facebook identity.  However, in other cases, the user may
  not have already revealed their identity to the site.  In general,
  IdPs SHOULD either verify that the user is willing to have their
  identity revealed to the site (e.g., through the usual IdP
  permissions dialog) or arrange that the identity information is only
  available to known RPs (e.g., social graph adjacencies) but not to
  the calling site.  The "domain" field of the assertion request can be
  used to check that the user has agreed to disclose their identity to
  the calling site; because it is supplied by the PeerConnection it can
  be trusted to be correct.

9.4.4.  Security of Third-Party IdPs

  As discussed above, each third-party IdP represents a new universal
  trust point and therefore the number of these IdPs needs to be quite
  limited.  Most IdPs, even those which issue unqualified identities
  such as Facebook, can be recast as authoritative IdPs (e.g.,
  [email protected]).  However, in such cases, the user interface
  implications are not entirely desirable.  One intermediate approach
  is to have special (potentially user configurable) UI for large
  authoritative IdPs, thus allowing the user to instantly grasp that
  the call is being authenticated by Facebook, Google, etc.

9.4.4.1.  Confusable Characters

  Because a broad range of characters are permitted in identity
  strings, it may be possible for attackers to craft identities which
  are confusable with other identities (see [RFC6943] for more on this
  topic).  This is a problem with any identifier space of this type
  (e.g., email addresses).  Those minting identifiers should avoid
  mixed scripts and similar confusable characters.  Those presenting
  these identifiers to a user should consider highlighting cases of
  mixed script usage (see [RFC5890], Section 4.4).  Other best
  practices are still in development.

9.4.5.  Web Security Feature Interactions

  A number of optional Web security features have the potential to
  cause issues for this mechanism, as discussed below.

9.4.5.1.  Popup Blocking

  When popup blocking is in use, the IdP proxy is unable to generate
  popup windows, dialogs, or any other form of user interactions.  This
  prevents the IdP proxy from being used to circumvent user
  interaction.  The "LOGINNEEDED" message allows the IdP proxy to
  inform the calling site of a need for user login, providing the
  information necessary to satisfy this requirement without resorting
  to direct user interaction from the IdP proxy itself.

9.4.5.2.  Third Party Cookies

  Some browsers allow users to block third party cookies (cookies
  associated with origins other than the top-level page) for privacy
  reasons.  Any IdP which uses cookies to persist logins will be broken
  by third-party cookie blocking.  One option is to accept this as a
  limitation; another is to have the PeerConnection object disable
  third-party cookie blocking for the IdP proxy.

10.  IANA Considerations

  This specification defines the "identity" SDP attribute per the
  procedures of Section 8.2.4 of [RFC4566].  The required information
  for the registration is included here:

  Contact Name:  IESG ([email protected])

  Attribute Name:  identity

  Long Form:  identity

  Type of Attribute:  session

  Charset Considerations:  This attribute is not subject to the charset
     attribute.

  Purpose:  This attribute carries an identity assertion, binding an
     identity to the transport-level security session.

  Appropriate Values:  See Section 5 of RFC 8827.

  Mux Category:  NORMAL

  This section registers the "idp-proxy" well-known URI from [RFC8615].

  URI suffix:  idp-proxy

  Change controller:  IETF

11.  References

11.1.  Normative References

  [FIPS186]  National Institute of Standards and Technology (NIST),
             "Digital Signature Standard (DSS)", NIST PUB 186-4,
             DOI 10.6028/NIST.FIPS.186-4, July 2013,
             <https://doi.org/10.6028/NIST.FIPS.186-4>.

  [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
             Requirement Levels", BCP 14, RFC 2119,
             DOI 10.17487/RFC2119, March 1997,
             <https://www.rfc-editor.org/info/rfc2119>.

  [RFC2818]  Rescorla, E., "HTTP Over TLS", RFC 2818,
             DOI 10.17487/RFC2818, May 2000,
             <https://www.rfc-editor.org/info/rfc2818>.

  [RFC3264]  Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model
             with Session Description Protocol (SDP)", RFC 3264,
             DOI 10.17487/RFC3264, June 2002,
             <https://www.rfc-editor.org/info/rfc3264>.

  [RFC3711]  Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K.
             Norrman, "The Secure Real-time Transport Protocol (SRTP)",
             RFC 3711, DOI 10.17487/RFC3711, March 2004,
             <https://www.rfc-editor.org/info/rfc3711>.

  [RFC3986]  Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform
             Resource Identifier (URI): Generic Syntax", STD 66,
             RFC 3986, DOI 10.17487/RFC3986, January 2005,
             <https://www.rfc-editor.org/info/rfc3986>.

  [RFC4566]  Handley, M., Jacobson, V., and C. Perkins, "SDP: Session
             Description Protocol", RFC 4566, DOI 10.17487/RFC4566,
             July 2006, <https://www.rfc-editor.org/info/rfc4566>.

  [RFC4568]  Andreasen, F., Baugher, M., and D. Wing, "Session
             Description Protocol (SDP) Security Descriptions for Media
             Streams", RFC 4568, DOI 10.17487/RFC4568, July 2006,
             <https://www.rfc-editor.org/info/rfc4568>.

  [RFC4648]  Josefsson, S., "The Base16, Base32, and Base64 Data
             Encodings", RFC 4648, DOI 10.17487/RFC4648, October 2006,
             <https://www.rfc-editor.org/info/rfc4648>.

  [RFC5763]  Fischl, J., Tschofenig, H., and E. Rescorla, "Framework
             for Establishing a Secure Real-time Transport Protocol
             (SRTP) Security Context Using Datagram Transport Layer
             Security (DTLS)", RFC 5763, DOI 10.17487/RFC5763, May
             2010, <https://www.rfc-editor.org/info/rfc5763>.

  [RFC5764]  McGrew, D. and E. Rescorla, "Datagram Transport Layer
             Security (DTLS) Extension to Establish Keys for the Secure
             Real-time Transport Protocol (SRTP)", RFC 5764,
             DOI 10.17487/RFC5764, May 2010,
             <https://www.rfc-editor.org/info/rfc5764>.

  [RFC5890]  Klensin, J., "Internationalized Domain Names for
             Applications (IDNA): Definitions and Document Framework",
             RFC 5890, DOI 10.17487/RFC5890, August 2010,
             <https://www.rfc-editor.org/info/rfc5890>.

  [RFC6347]  Rescorla, E. and N. Modadugu, "Datagram Transport Layer
             Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347,
             January 2012, <https://www.rfc-editor.org/info/rfc6347>.

  [RFC6454]  Barth, A., "The Web Origin Concept", RFC 6454,
             DOI 10.17487/RFC6454, December 2011,
             <https://www.rfc-editor.org/info/rfc6454>.

  [RFC7022]  Begen, A., Perkins, C., Wing, D., and E. Rescorla,
             "Guidelines for Choosing RTP Control Protocol (RTCP)
             Canonical Names (CNAMEs)", RFC 7022, DOI 10.17487/RFC7022,
             September 2013, <https://www.rfc-editor.org/info/rfc7022>.

  [RFC7675]  Perumal, M., Wing, D., Ravindranath, R., Reddy, T., and M.
             Thomson, "Session Traversal Utilities for NAT (STUN) Usage
             for Consent Freshness", RFC 7675, DOI 10.17487/RFC7675,
             October 2015, <https://www.rfc-editor.org/info/rfc7675>.

  [RFC7918]  Langley, A., Modadugu, N., and B. Moeller, "Transport
             Layer Security (TLS) False Start", RFC 7918,
             DOI 10.17487/RFC7918, August 2016,
             <https://www.rfc-editor.org/info/rfc7918>.

  [RFC8122]  Lennox, J. and C. Holmberg, "Connection-Oriented Media
             Transport over the Transport Layer Security (TLS) Protocol
             in the Session Description Protocol (SDP)", RFC 8122,
             DOI 10.17487/RFC8122, March 2017,
             <https://www.rfc-editor.org/info/rfc8122>.

  [RFC8174]  Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
             2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
             May 2017, <https://www.rfc-editor.org/info/rfc8174>.

  [RFC8259]  Bray, T., Ed., "The JavaScript Object Notation (JSON) Data
             Interchange Format", STD 90, RFC 8259,
             DOI 10.17487/RFC8259, December 2017,
             <https://www.rfc-editor.org/info/rfc8259>.

  [RFC8261]  Tuexen, M., Stewart, R., Jesup, R., and S. Loreto,
             "Datagram Transport Layer Security (DTLS) Encapsulation of
             SCTP Packets", RFC 8261, DOI 10.17487/RFC8261, November
             2017, <https://www.rfc-editor.org/info/rfc8261>.

  [RFC8445]  Keranen, A., Holmberg, C., and J. Rosenberg, "Interactive
             Connectivity Establishment (ICE): A Protocol for Network
             Address Translator (NAT) Traversal", RFC 8445,
             DOI 10.17487/RFC8445, July 2018,
             <https://www.rfc-editor.org/info/rfc8445>.

  [RFC8446]  Rescorla, E., "The Transport Layer Security (TLS) Protocol
             Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018,
             <https://www.rfc-editor.org/info/rfc8446>.

  [RFC8615]  Nottingham, M., "Well-Known Uniform Resource Identifiers
             (URIs)", RFC 8615, DOI 10.17487/RFC8615, May 2019,
             <https://www.rfc-editor.org/info/rfc8615>.

  [RFC8825]  Alvestrand, H., "Overview: Real-Time Protocols for
             Browser-Based Applications", RFC 8825,
             DOI 10.17487/RFC8825, January 2021,
             <https://www.rfc-editor.org/info/rfc8825>.

  [RFC8826]  Rescorla, E., "Security Considerations for WebRTC",
             RFC 8826, DOI 10.17487/RFC8826, January 2021,
             <https://www.rfc-editor.org/info/rfc8826>.

  [RFC8829]  Uberti, J., Jennings, C., and E. Rescorla, Ed.,
             "JavaScript Session Establishment Protocol (JSEP)",
             RFC 8829, DOI 10.17487/RFC8829, January 2021,
             <https://www.rfc-editor.org/info/rfc8829>.

  [RFC8834]  Perkins, C., Westerlund, M., and J. Ott, "Media Transport
             and Use of RTP in WebRTC", RFC 8834, DOI 10.17487/RFC8834,
             January 2021, <https://www.rfc-editor.org/info/rfc8834>.

  [RFC8844]  Thomson, M. and E. Rescorla, "Unknown Key-Share Attacks on
             Uses of TLS with the Session Description Protocol (SDP)",
             RFC 8844, DOI 10.17487/RFC8844, January 2021,
             <https://www.rfc-editor.org/info/rfc8844>.

  [webcrypto]
             Watson, M., "Web Cryptography API", W3C Recommendation, 26
             January 2017,
             <https://www.w3.org/TR/2017/REC-WebCryptoAPI-20170126/>.

  [webrtc-api]
             Jennings, C., Boström, H., and J-I. Bruaroey, "WebRTC 1.0:
             Real-time Communication Between Browsers", W3C Proposed
             Recommendation, <https://www.w3.org/TR/webrtc/>.

11.2.  Informative References

  [fetch]    van Kesteren, A., "Fetch",
             <https://fetch.spec.whatwg.org/>.

  [RFC3261]  Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston,
             A., Peterson, J., Sparks, R., Handley, M., and E.
             Schooler, "SIP: Session Initiation Protocol", RFC 3261,
             DOI 10.17487/RFC3261, June 2002,
             <https://www.rfc-editor.org/info/rfc3261>.

  [RFC5705]  Rescorla, E., "Keying Material Exporters for Transport
             Layer Security (TLS)", RFC 5705, DOI 10.17487/RFC5705,
             March 2010, <https://www.rfc-editor.org/info/rfc5705>.

  [RFC6120]  Saint-Andre, P., "Extensible Messaging and Presence
             Protocol (XMPP): Core", RFC 6120, DOI 10.17487/RFC6120,
             March 2011, <https://www.rfc-editor.org/info/rfc6120>.

  [RFC6265]  Barth, A., "HTTP State Management Mechanism", RFC 6265,
             DOI 10.17487/RFC6265, April 2011,
             <https://www.rfc-editor.org/info/rfc6265>.

  [RFC6455]  Fette, I. and A. Melnikov, "The WebSocket Protocol",
             RFC 6455, DOI 10.17487/RFC6455, December 2011,
             <https://www.rfc-editor.org/info/rfc6455>.

  [RFC6943]  Thaler, D., Ed., "Issues in Identifier Comparison for
             Security Purposes", RFC 6943, DOI 10.17487/RFC6943, May
             2013, <https://www.rfc-editor.org/info/rfc6943>.

  [RFC7617]  Reschke, J., "The 'Basic' HTTP Authentication Scheme",
             RFC 7617, DOI 10.17487/RFC7617, September 2015,
             <https://www.rfc-editor.org/info/rfc7617>.

  [RFC8224]  Peterson, J., Jennings, C., Rescorla, E., and C. Wendt,
             "Authenticated Identity Management in the Session
             Initiation Protocol (SIP)", RFC 8224,
             DOI 10.17487/RFC8224, February 2018,
             <https://www.rfc-editor.org/info/rfc8224>.

  [RFC8828]  Uberti, J. and G. Shieh, "WebRTC IP Address Handling
             Requirements", RFC 8828, DOI 10.17487/RFC8828, January
             2021, <https://www.rfc-editor.org/info/rfc8828>.

  [TLS-DTLS13]
             Rescorla, E., Tschofenig, H., and N. Modadugu, "The
             Datagram Transport Layer Security (DTLS) Protocol Version
             1.3", Work in Progress, Internet-Draft, draft-ietf-tls-
             dtls13-39, 2 November 2020,
             <https://tools.ietf.org/html/draft-ietf-tls-dtls13-39>.

Acknowledgements

  Bernard Aboba, Harald Alvestrand, Richard Barnes, Dan Druta, Cullen
  Jennings, Hadriel Kaplan, Matthew Kaufman, Jim McEachern, Martin
  Thomson, Magnus Westerlund.  Matthew Kaufman provided the UI material
  in Section 6.5.  Christer Holmberg provided the initial version of
  Section 5.1.

Author's Address

  Eric Rescorla
  Mozilla

  Email: [email protected]

You are viewing proxied material from gopher.fnord.one. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.