[Docs] [txt|pdf] [Tracker] [Email] [Nits]

Versions: 00 01

ecrit                                                           B. Rosen
Internet-Draft                                                   NeuStar
Expires: August 31, 2006                                         J. Polk
                                                           Cisco Systems
                                                       February 27, 2006

    Best Current Practice for Communications Services in support of
                           Emergency Calling

Status of this Memo

   By submitting this Internet-Draft, each author represents that any
   applicable patent or other IPR claims of which he or she is aware
   have been or will be disclosed, and any of which he or she becomes
   aware will be disclosed, in accordance with Section 6 of BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   The list of current Internet-Drafts can be accessed at

   The list of Internet-Draft Shadow Directories can be accessed at

   This Internet-Draft will expire on August 31, 2006.

Copyright Notice

   Copyright (C) The Internet Society (2006).


   Requesting help in an emergency using a communications device such as
   a telephone or mobile is an accepted practice in most of the world.
   As communications devices increasingly utilize the Internet to
   interconnect and communicate, users will continue to expect to use
   such devices to request help, regardless of whether or not they
   communicate using IP.  The emergency response community will have to

Rosen & Polk             Expires August 31, 2006                [Page 1]

Internet-Draft          Emergency Call Phone BCP           February 2006

   upgrade their facilities to support the wider range of communications
   services, but cannot be expected to handle wide variation in device
   and service capability.  The IETF has several efforts targeted at
   standardizing various aspects of placing emergency calls.  This memo
   describes best current practice on how devices and services should
   use such standards to reliably make emergency calls

Table of Contents

   1.  Requirements notation  . . . . . . . . . . . . . . . . . . . .  3
   2.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  3
   3.  Which devices and services should support emergency calls  . .  4
   4.  Determining Location . . . . . . . . . . . . . . . . . . . . .  4
   5.  Determining an emergency call  . . . . . . . . . . . . . . . .  6
   6.  Session Signaling  . . . . . . . . . . . . . . . . . . . . . .  7
     6.1.  SIP signaling requirements for User Agents . . . . . . . .  8
     6.2.  Mapping from Location to a PSAP URI  . . . . . . . . . . .  9
     6.3.  Routing the call . . . . . . . . . . . . . . . . . . . . .  9
     6.4.  Responding to PSAP signaling . . . . . . . . . . . . . . . 10
     6.5.  Disabling of features  . . . . . . . . . . . . . . . . . . 10
   7.  Security Considerations  . . . . . . . . . . . . . . . . . . . 11
   8.  Normative References . . . . . . . . . . . . . . . . . . . . . 11
   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 13
   Intellectual Property and Copyright Statements . . . . . . . . . . 14

Rosen & Polk             Expires August 31, 2006                [Page 2]

Internet-Draft          Emergency Call Phone BCP           February 2006

1.  Requirements notation

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   document are to be interpreted as described in [RFC2119].

2.  Introduction

   In this memo, an emergency call refers to a communications session
   established by a user to a "Public Safety Answering Point" (PSAP)
   which is a call center established by response agencies to accept
   emergency calls.  We differentiate such calls from other sessions
   which are created by responders using public communications
   infrastructure often involving some kind of priority access as
   defined in Emergency Telecommunications Service (ETS) in IP Telephony
   (RFC4190).  While current PSAPs are limited to voice sessions, often
   with the additional capability to serve hearing impaired users with
   text based "TTY" devices, envisioned upgrades to PSAPs will allow
   sessions with audio, video, and several kinds of text including
   interactive text (RFC4103) and Instant Messages.

   Making an emergency call involves the use of location information,
   referring to the physical location of the caller.  Location is used
   within the emergency calling system to route a call to the correct
   PSAP, as well as by the PSAP to choose the correct responder, and
   direct them to the person in need of assistance.

   The steps involved in an emergency call are (with a rough ordering of
   1.  User dials visited location's emergency number
   2.  User device identifies call as emergency call
   3.  User device includes location indication in the call set-up
   4.  emergency call set-up is routed to appropriate PSAP based on
       location of the caller
   5.  call is established with PSAP
   6.  caller's location is presented to PSAP operator for dispatch

   As a quick overview for a typical Ethernet connected telephone using
   SIP signaling:
   o  the phone would get location from the DHCP server [or an L7
      server] when it boots.
   o  It would use "urn:service:sos" as the URI of an emergency call.
   o  It would put its location in the SIP INVITE as a PIDF-LO in the
      body of the INVITE (or a reference to location in a Location
      header) and forward the call to its first hop proxy.

Rosen & Polk             Expires August 31, 2006                [Page 3]

Internet-Draft          Emergency Call Phone BCP           February 2006

   o  The proxy recognize the call as an emergency call.
   o  The proxy would determine the PSAP's URI by using the [ecrit]
      mapping server from the location provided in the signaling
   o  The proxy would use a SIP SRV record in the domain of the
      resulting PSAP URI to determine where to send the call.

   The (upgraded) PSAP would answer the call as SIP, with location

3.  Which devices and services should support emergency calls

   Although present PSAPs have only support for voice calls placed
   through PSTN facilities or systems connected to the PSTN, future
   PSAPs will support Internet connectivity and a wider range of media
   types.  In general, if a user could reasonably expect to be able to
   call for help with the device, then the device or service should
   support emergency calling.  Certainly, any device or service that
   looks like and works like a telephone (wired or mobile) should
   support emergency calling, but increasingly, users have expectations
   that other devices and services should work.

   Using current (evolving) standards, devices that create media
   sessions and exchange audio, video and/or text, and have the
   capability to establish sessions to a wide variety of addresses, and
   communicate over private IP networks or the Internet, should support
   emergency calls.

4.  Determining Location

   With Internet based communications services, determining where the
   caller is located is more problematic than in PSTN and mobile
   systems.  Existing wired phones are tethered with a wire that is
   connected directly to a call control device, a circuit switch.
   Cellular phones are tethered via a radio channel to a cell tower,
   which connects that cell phone to a circuit switch.  The primary
   difficulty with IP based phones is that the connectivity, whether
   wired or radio channel, is decoupled from the call control device.
   The communications service may not have any relationship with the
   access network carrier, and, with NAT and VPN tunnels, may have no
   way to even find out who the access carrier is.

   For this reason, standards have been created for endpoints (devices)
   to obtain location information.  The endpoint is a subscriber to both
   the access network and the communications service, and thus is in a
   position to obtain its location from the access network, and supply
   it to the communications service.

Rosen & Polk             Expires August 31, 2006                [Page 4]

Internet-Draft          Emergency Call Phone BCP           February 2006

   DHCP (RFC2131) has been enhanced to provide the location of a device.
   RFC3825 describes how a geo-location (lat/lon/alt) may be obtained
   and draft-ietf-geopriv-dhcp-civil-09 describes how a civic (street
   address) location can be obtained via DHCP.

   [Placeholder for HELD, LCP or other L7 location determination

   For devices that operate on a network where the network operator
   controls the specification of every device connected to that network
   that could be used for emergency calls, the method by which location
   is determined need not be an IETF standard, but can be any method
   that achieves the desired result.  Such a method MUST be specified,
   and every device MUST support it.

   For devices that operate in a network where the network operator
   controls the specification of every device connected to that network,
   but the network attachment supports upstream networks to which
   communications devices are connected (such as any network that
   supports Ethernet connected telephones and terminal adapters), the
   method by which location is determined need not be an IETF standard,
   but can be any method which achieves the desired result.  However,
   the network attachment must support [both] DHCP [AND L7] for upstream
   communications devices to obtain location.  For smaller interior
   (e.g, LAN) networks, the DHCP [or L7] server should simply repeat the
   location obtained from the access network.  For larger networks,
   other mechanisms, such as a DHCP Relay Agent (RFC3046) must be used
   to provide more accurate location of endpoints.

   For devices that operate on a network where the network operator does
   not control the specification of every device connected to the
   network, DHCP [or L7] MUST be supported on the network.

   Note: Self Reported location is generally unacceptable in emergency
   calls, although it is being used prior to automatic location
   determination schemes being fielded.  Local laws may govern what is
   acceptable in any country or area.

   Devices should get location immediately after obtaining local network
   configuration information.  It is essential for the location to be
   determined BEFORE any VPN tunnels are established.  It is equally
   essential that this location information is *not* overwritten by any
   process engaged from establishing a VPN connection.  In other words,
   the established VPN to Chicago from the device in Dallas should not
   overwrite the location of "Dallas".

   It is desirable that location information be periodically refreshed.
   For devices which are not expected to roam, refreshing on the order

Rosen & Polk             Expires August 31, 2006                [Page 5]

Internet-Draft          Emergency Call Phone BCP           February 2006

   of once per day is recommended.  For devices which roam, refresh of
   location should be more frequent, with the frequency related to the
   mobility of the device.  There can be instances in which a device is
   aware of when it moves, for example when it changes access points.
   When this type of event occurs, the device should refresh its

   It is desirable for location information to be requested immediately
   before placing an emergency call.  However, if there is any delay in
   getting more recent location, the call should be placed with the most
   recent location information the device has.  It is recommended that
   the device not wait longer than 500 ms to obtain updated location,
   and systems should be designed such that the typical response is
   under 100ms.

5.  Determining an emergency call

   An emergency call is distinguished by the device (or a downstream
   element) by an "address", which in most cases for Internet connected
   devices is still a dialstring, although other user interfaces may be

   Note: It is undesirable to have a single "button" emergency call user
   interface element.  These mechanisms have a very high false call
   rate.  PSAPs prefer devices to use the local emergency call

   While in some countries there is a single 3 digit dialstring that is
   used for all emergency calls (i.e. 911 in North America), in some
   countries there are several 3 digit numbers used for different types
   of calls.  For example, in Switzerland, 117 is used to call police,
   118 is used to call the fire brigade, and 144 is used for emergency
   medical assistance.  In other countries, there are no "short codes"
   or "service codes" for 3 digit dialing of emergency services and
   local (PSTN) numbers are used.

   draft-schulzrinne-sipping-service-01.txt introduces a universal
   emergency service urn scheme.  On the wire, emergency calls should
   include this type of URI (in for example, the To: field of a SIP
   call).  The scheme includes a single emergency URN (urn:service:sos)
   and responder specific ones (urn:service:sos.police).  Using the
   service sos urn scheme, emergency calls can be recognized as such
   throughout the Internet.

   Devices MUST use the service:sos urn scheme to mark emergency calls.

Rosen & Polk             Expires August 31, 2006                [Page 6]

Internet-Draft          Emergency Call Phone BCP           February 2006

   To determine which calls are emergency calls, some entity needs to
   map a user entered dialstring into this URN scheme.  A user may
   "dial" 1-1-2, but the call would be sent to urn:service:sos.  This
   mapping is ideally performed at the endpoint device, but may be
   performed at an intermediate entity (such as a SIP proxy server).

   Note: It is strongly suggested that devices recognize the emergency
   dialstring(s) and map to the universal emergency urn.  If devices
   cannot do "dial plan interpretation", then the first signaling aware
   element (first hop proxy in SIP signaled devices) should do the
   mapping.  It is important to not require a large number of active
   elements handle a call before it is recognized as an emergency call

   In systems that support roaming, there may be a concept of "visited"
   and "home" networks.  Even when there is not a "visited network", the
   user may be roaming (or nomadic) in a different country from their
   home.  This gives rise to the problem of which dialstring(s) to
   recognize, the "home" or "visited"?  While it is desirable that the
   "home" dialstrings be recognized, it is required (by law in some
   countries) that the "visited" dialstrings be recognized.  Dial plan
   interpretation may need to take "visited" emergency dialstrings into

   To give an example of this difference in dialstrings: If the device
   is from North America, the home and visited emergency dialstring is
   "9-1-1".  If that devices roams to the UK, the home emergency
   dialstring is still "9-1-1", but the visited emergency dialstring
   would become "9-9-9".  If the device roams to London, the home
   dialstring remains the same, "9-1-1", but the visited dialstring
   changes from 999 to "1-1-2".

   The home emergency dialstrings may be provisioned into the device (or
   other element doing dialstring to universal emergency call urn
   mapping).  The visited dialstring may be discovered by a lower layer
   protocol that is used by the access network, such as DHCP, or with a
   higher layer protocol like SIP (using a REGISTER Request) or HTTP
   (using a GET Request) once the device learns its location.  It could
   be that the device knows more than one way to learn the visited
   emergency dialstring, and using the methods in some configured order
   (until an answer is received).

6.  Session Signaling

   SIP signaling (RFC3261) is expected be supported by upgraded PSAPs.
   Gateways may be used between Internet connected devices and older
   PSAPs.  Some countries may support other signaling protocols into

Rosen & Polk             Expires August 31, 2006                [Page 7]

Internet-Draft          Emergency Call Phone BCP           February 2006

6.1.  SIP signaling requirements for User Agents

   Initial signaling Method is INVITE.  The Request URI must be a
   service:sos URN unless the device does not do emergency dialstring
   interpretation.  If the device does not do emergency dialstring
   interpretation, the expectation is that the request URI will be a tel
   URI with the dialed digits, or a sips uri with the dialed digits and
   a USER=PHONE parameter (e.g. sips:911@example.com;user=phone).  The
   call would normally be sent to the first hop proxy of the
   communications service.
   1.   The To: header must be present and should be the same as the
        Request URI
   2.   The From: header must be present and should be the AoR of the
   3.   NOTE: this may be the contact address of this device to ensure
        callback is to this specific device
   4.   A Via: header must be present and should include the URI of the
   5.   A Route header MAY be present if the device has performed a
        fallback mapping function (see section 4)
   6.   Other headers MAY be included as per normal sip behavior
   7.   A Supported: header MUST be included with the 'location' option
        tag, unless the device does not understand the concept of SIP
        Location ;
   8.   If the device's location is by-reference, a Location: header
        MUST be present containing the URI of the PIDF-LO reference for
        that device;
   9.   If the device does not know its location, but understands the
        concept of SIP Location, the Location: header MUST be present
        with an unknown option tag;
   10.  If the device knows its location and chooses to include this by-
        value in the INVITE message, the request MUST contain a PIDF-LO
        message body part [RFC4119], which implies Content-Type: header
        value of MIME/Multipart
   11.  A normal SDP offer SHOULD be included in the INVITE.  The offer
        SHOULD NOT include compressed audio codecs, although a wideband
        codec offer may be included.

   Note: Silence suppression (Voice Activity Detection methods) MUST NOT
   be used on emergency calls.  PSAP call takers sometimes get
   information on what is happening in the background to determine how
   to process the call.

   It is RECOMMENDED that location by-value is used by the device to
   convey location during emergency calling.

Rosen & Polk             Expires August 31, 2006                [Page 8]

Internet-Draft          Emergency Call Phone BCP           February 2006

6.2.  Mapping from Location to a PSAP URI

   To route an emergency call, we make use of the [ecrit] mapping
   service which takes a location expressed by a PIDF-LO and returns one
   or more PSAP URIs.  The request includes the service urn which is
   used to determine which entity should receive the call.  The URI
   would replace the Request URI in a SIP INVITE.

   It is desirable for the endpoint to do the mapping from location to
   PSAP URI.  This will allow for a fallback map to be know.  It is
   desirable for the first active signaling element (first hop proxy for
   SIP) to do an [ecrit] mapping as well, as this will be the freshest
   and most up-to-date URI of the appropriate PSAP.  If the device
   included a PSAP URI from fallback mapping in the SIP INVITE, the
   first SIP element SHOULD do its mapping and use this result to
   forward the message towards the PSAP.  The fallback mapping's purpose
   is in case the SIP element's mapping failed, the element would have a
   route already in the message, just not one that is the freshest.

   At the time of a call, it may be the case that the element doing the
   mapping cannot reach the mapping server, or the server could be
   congested.  To guard against the inability to correctly route, or to
   avoid undue delay, it is desirable that an older mapping be kept
   (cached).  The device, if it is doing the mapping, should request a
   mapping whenever it gets location and retain that mapping for use in
   unusual circumstances.  If the first active element maps, it has the
   problem that it will not get location for the device until an
   emergency call is placed.  Thus the first hop proxy cannot do a
   mapping in advance.

   In this case, the DHCP server [or L7 server] will return the mapping
   for the supplied location, which can be retained by the device.  When
   an emergency call is placed, it sends this cached mapping in a
   (loose) Route header.

   Mapping elements should start a timer when requesting a mapping (TLS
   establishment if appropriate), with a time-out of 500ms.  If the
   timer expires before the mapping is complete, the cached mapping
   should be used.

6.3.  Routing the call

   Normal routing mechanisms for the specified URI should be used.  For
   SIP signaled devices, the domain of the URI should be extracted, and
   the DNS consulted for a sip (or sips) SRV.  The resulting NAPTR, if
   present, should be used for the FQDN of the server.

Rosen & Polk             Expires August 31, 2006                [Page 9]

Internet-Draft          Emergency Call Phone BCP           February 2006

6.4.  Responding to PSAP signaling

   The PSAP is expected to use normal signaling (e.g.  SIP) as per IETF
   standards.  Devices and proxies should expect to:
   1.  Be REFERed to a conference bridge; PSAPs often include
       dispatchers, responders or specialists on a call.
   2.  Be REFERed to a secondary PSAP.  Some responder's dispatchers are
       not located in the primary PSAP.  The call may have to be
       transferred to another PSAP.  Most often this will be an attended
       transfer, or a bridged transfer.
   3.  (For devices that are Mobile) SUBSCRIBE to the Presence of the
       AoR (or equivalent for other signaling schemes) to get location
   4.  Support Session Timer (or equivalent) to guard against session

   Devices MUST NOT send a BYE (or equivalent for other non-SIP
   signaling).  The PSAP must be the only entity that can terminate a
   call.  If the user "hangs up" an emergency call, the device should
   ring, and when answered, reconnect the caller to the PSAP.

   There can be a case where the session signaling path is lost, and the
   user agent does not receive the BYE.  If the call is hung up, the
   session timer expires, and 5 minutes elapses from the last message
   received by the device from the PSAP, the call may be declared lost.
   If in the 5 minute interval an incoming call is received from the
   domain of the PSAP, the device should drop the old call and alert for
   the (new) incoming call.

6.5.  Disabling of features

   The device and/or service should disable outgoing call features such
   o  Call Waiting
   o  Call Transfer
   o  Three Way Call
   o  Flash hold
   o  Outbound Call Blocking

   The emergency dialstrings should not be permitted in Call Forward
   numbers or speed dial lists.

   The device and/or service should disable the following incoming call
   features on calls from the PSAP:
   o  Call Waiting (all kinds)
   o  Do Not Disturb

Rosen & Polk             Expires August 31, 2006               [Page 10]

Internet-Draft          Emergency Call Phone BCP           February 2006

   o  Call Forward (all kinds) (if the PSAP calls back within some
      (30min?) interval)

7.  Security Considerations

   There are no new security considerations beyond those in the
   normative references.  This memo does not introduce any new
   protocols; it specifies use of several of them.  Implementors are
   admonished to ,,,

8.  Normative References

              Polk, J. and B. Rosen, "Session Initiation Protocol
              Location Conveyance",
              draft-ietf-sip-location-conveyance-01 (work in progress),
              July 2005.

              Wijk, A., "Framework of requirements for real-time text
              conversation using SIP", draft-ietf-sipping-toip-03 (work
              in progress), September 2005.

              Schulzrinne, H., "DHCP Option for Civil Location",
              draft-schulzrinne-geopriv-dhcp-civil-01 (work in
              progress), February 2003.

              Schulzrinne, H., "A Uniform Resource Name (URN) for
              Services", draft-schulzrinne-sipping-service-01 (work in
              progress), October 2005.

              Sinnreich, H., "SIP Telephony Device Requirements and
              Configuration", draft-sinnreich-sipdev-req-08 (work in
              progress), October 2005.

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119, March 1997.

   [RFC2131]  Droms, R., "Dynamic Host Configuration Protocol",
              RFC 2131, March 1997.

   [RFC3046]  Patrick, M., "DHCP Relay Agent Information Option",
              RFC 3046, January 2001.

Rosen & Polk             Expires August 31, 2006               [Page 11]

Internet-Draft          Emergency Call Phone BCP           February 2006

   [RFC3261]  Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston,
              A., Peterson, J., Sparks, R., Handley, M., and E.
              Schooler, "SIP: Session Initiation Protocol", RFC 3261,
              June 2002.

   [RFC3825]  Polk, J., Schnizlein, J., and M. Linsner, "Dynamic Host
              Configuration Protocol Option for Coordinate-based
              Location Configuration Information", RFC 3825, July 2004.

   [RFC4103]  Hellstrom, G. and P. Jones, "RTP Payload for Text
              Conversation", RFC 4103, June 2005.

   [RFC4119]  Peterson, J., "A Presence-based GEOPRIV Location Object
              Format", RFC 4119, December 2005.

   [RFC4190]  Carlberg, K., Brown, I., and C. Beard, "Framework for
              Supporting Emergency Telecommunications Service (ETS) in
              IP Telephony", RFC 4190, November 2005.

Rosen & Polk             Expires August 31, 2006               [Page 12]

Internet-Draft          Emergency Call Phone BCP           February 2006

Authors' Addresses

   Brian Rosen
   470 Conrad Dr.
   Mars, PA  16046

   Phone: +1 724 382 1051
   Email: br@brianrosen.net

   James M. Polk
   Cisco Systems
   3913 Treemont Circle
   Colleyville, TX  76034

   Phone: +1-817-271-3552
   Email: jmpolk@cisco.com

Rosen & Polk             Expires August 31, 2006               [Page 13]

Internet-Draft          Emergency Call Phone BCP           February 2006

Intellectual Property Statement

   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights.  Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard.  Please address the information to the IETF at

Disclaimer of Validity

   This document and the information contained herein are provided on an

Copyright Statement

   Copyright (C) The Internet Society (2006).  This document is subject
   to the rights, licenses and restrictions contained in BCP 78, and
   except as set forth therein, the authors retain all their rights.


   Funding for the RFC Editor function is currently provided by the
   Internet Society.

Rosen & Polk             Expires August 31, 2006               [Page 14]

Html markup produced by rfcmarkup 1.129b, available from https://tools.ietf.org/tools/rfcmarkup/