Rfc | 3603 |
Title | Private Session Initiation Protocol (SIP) Proxy-to-Proxy Extensions
for Supporting the PacketCable Distributed Call Signaling
Architecture |
Author | W. Marshall, Ed., F. Andreasen, Ed. |
Date | October 2003 |
Format: | TXT, HTML |
Obsoleted by | RFC5503 |
Status: | INFORMATIONAL |
|
Network Working Group W. Marshall, Ed.
Request for Comments: 3603 AT&T
Category: Informational F. Andreasen, Ed.
Cisco
October 2003
Private Session Initiation Protocol (SIP) Proxy-to-Proxy Extensions
for Supporting the PacketCable Distributed Call Signaling Architecture
Status of this Memo
This memo provides information for the Internet community. It does
not specify an Internet standard of any kind. Distribution of this
memo is unlimited.
Copyright Notice
Copyright (C) The Internet Society (2003). All Rights Reserved.
Abstract
In order to deploy a residential telephone service at very large
scale across different domains, it is necessary for trusted elements
owned by different service providers to exchange trusted information
that conveys customer-specific information and expectations about the
parties involved in the call. This document describes private
extensions to the Session Initiation Protocol (SIP) (RFC3261) for
supporting the exchange of customer information and billing
information between trusted entities in the PacketCable Distributed
Call Signaling Architecture. These extensions provide mechanisms for
access network coordination to prevent theft of service, customer
originated trace of harassing calls, support for operator services
and emergency services, and support for various other regulatory
issues. The use of the extensions is only applicable within closed
administrative domains, or among federations of administrative
domains with previously agreed-upon policies where coordination of
charging and other functions is required.
Table of Contents
1. Applicability Statement . . . . . . . . . . . . . . . . . . . 3
2. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . 3
3. Trust Boundary. . . . . . . . . . . . . . . . . . . . . . . . 5
4. Conventions used in this document . . . . . . . . . . . . . . 6
1. Applicability Statement
The SIP extensions described in this document make certain
assumptions regarding network topology, linkage between SIP and lower
layers, and the availability of transitive trust. These assumptions
are generally not applicable in the Internet as a whole. The use of
these headers is only applicable within closed administrative
domains, or among federations of administrative domains with
previously agreed-upon policies where coordination of charging and
other functions is required, as in for example the architecture
presented in [6]. Use outside such a domain could result in the
leakage of potentially sensitive or private information. User
consent to the privacy implications of the policies in [6] is
strongly encouraged in those domains as well.
Although RFC 2119 language is used in this document, the scope of the
normative language is only for the area of applicability of the
document and, like the technology, it does not apply to the general
Internet.
2. Introduction
In order to deploy a SIP-based [2] residential telephone service at
very large scale across different domains, it is necessary for
trusted elements owned by different service providers to exchange
trusted information that conveys billing information and expectations
about the parties involved in the call.
There are many billing models used in deriving revenue from telephony
services today. Charging for telephony services is tightly coupled
to the use of network resources. It is outside the scope of this
document to discuss the details of these numerous and varying
methods.
A key motivating principle of the DCS architecture described in [6]
is the need for network service providers to be able to control and
monitor network resources; revenue may be derived from the usage of
these resources as well as from the delivery of enhanced services
such as telephony. Furthermore, the DCS architecture recognizes the
need for coordination between call signaling and resource management.
This coordination ensures that users are authenticated and authorized
before receiving access to network resources and billable enhanced
services.
DCS Proxies, as defined in [6], have access to subscriber information
and act as policy decision points and trusted intermediaries along
the call signaling path. Edge routers provide the network
connectivity and resource policy enforcement mechanism and also
capture and report network connectivity and resource usage
information. Edge routers need to be given billing information that
can be logged with Record Keeping or Billing servers. The DCS Proxy,
as a central point of coordination between call signaling and
resource management, can provide this information based on the
authenticated identity of the calling and called parties. Since
there is a trust relationship among DCS Proxies, they can be relied
upon to exchange trusted billing information pertaining to the
parties involved in a call. See [6] for a description of the trust
boundary and trusted versus untrusted entities.
For these reasons, it is appropriate to consider defining SIP header
extensions to allow DCS Proxies to exchange information during call
setup. It is the intent that the extensions would only appear on
trusted network segments, should be inserted upon entering a trusted
network region, and removed before leaving trusted network segments.
Significant amounts of information is retrieved by an originating DCS
Proxy in its handling of a connection setup request from a user
agent. Such information includes location information about the
subscriber (essential for emergency services calls), billing
information, and station information (e.g., coin operated phone). In
addition, while translating the destination number, information such
as the local-number-portability office code is obtained and will be
needed by all other proxies handling this call.
For Usage Accounting records, it is necessary to have an identifier
that can be associated with all the event records produced for the
call. The SIP Call-ID header field cannot be used as such an
identifier since it is selected by the originating user agent, and
may not be unique among all past calls as well as current calls.
Further, since this identifier is to be used by the service provider,
it should be chosen in a manner and in a format that meets the
service provider's needs.
Billing information may not necessarily be unique for each user
(consider the case of calls from an office all billed to the same
account). Billing information may not necessarily be identical for
all calls made by a single user (consider prepaid calls, credit card
calls, collect calls, etc). It is therefore necessary to carry
billing information separate from the calling and called party
identification. Furthermore, some billing models call for split-
charging where multiple entities are billed for portions of the call.
The addition of a SIP General Header Field allows for the capture of
billing information and billing identification for the duration of
the call.
It is the intent that the billing extensions would only appear on
trusted network segments, and MAY be inserted by a DCS Proxy in
INVITE and REFER requests and INVITE responses in a trusted network
segment, and removed before leaving trusted network segments.
In addition to support for billing, current residential telephone
service includes the need for customer originated trace (of harassing
or obscene calls), for operator services such as busy line
verification and emergency interrupt (initiated by an operator from
an Operator Services Position System (OSPS)), for emergency services
such as 9-1-1 calls to a Public Service Access Point (PSAP) and the
subsequent call handling, and support for Electronic Surveillance and
Law Enforcement access as required by applicable legislation and
court orders. In all of these cases, additional information about
the call and about the subscribers involved in the call needs to be
exchanged between the proxies.
3. Trust Boundary
The DCS architecture [6] defines a trust boundary around the various
systems and servers that are owned, operated by, and/or controlled by
the service provider. These trusted systems include the proxies and
various servers such as bridge servers, voicemail servers,
announcement servers, etc. Outside of the trust boundary lie the
customer premises equipment, and various application and media
servers operated by third-party service providers.
Certain subscriber-specific information, such as billing and
accounting information, stays within the trust boundary. Other
subscriber-specific information, such as endpoint identity, may be
presented to untrusted endpoints or may be withheld based on
subscriber profiles.
The User Agent (UA) may be either within the trust boundary or
outside the trust boundary, depending on exactly what function is
being performed and exactly how it is being performed. Accordingly,
the procedures followed by a User Agent are different depending on
whether the UA is within the trust boundary or outside the trust
boundary.
The following sections giving procedures for User Agents therefore
are subdivided into trusted user agents and untrusted user agents.
4. Conventions used in this document
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in BCP 14, RFC 2119 [1].
The term "private-URL" used in this document refers to a SIP URI that
is generated by a proxy, contains a "hostport" that identifies the
proxy, and contains a "userinfo" string that is generated by the
proxy. The "userinfo" typically contains (or points to) information
that is not to be disclosed outside the trusted domain of the
proxies, such as billing account numbers, electronic surveillance
indication, electronic surveillance parameters, and call redirection
information. Consequently, the information is either stored locally
by the proxy, or encrypted with a private key known only to the proxy
and encoded in a character string in the "userinfo" portion of the
URL. A checksum is included in the "userinfo" data to detect
tampering. The mechanism by which a proxy recognizes a "userinfo" as
a private-URL and decodes and recovers the original information is
local to the proxy and is not subject to standardization. Some
possible implementations include an initial magic cookie (e.g.,
z9hG4Bk followed by the pointer/information), or use of a reserved
"user" name (e.g., "private") with the optional "password" containing
the pointer/information.
5. P-DCS-TRACE-PARTY-ID
In the telephone network, calling identity information is used to
support regulatory requirements such as the Customer Originated Trace
service, which provide the called party with the ability to report
obscene or harassing phone calls to law enforcement. This service is
provided independently of caller-id, and works even if the caller
requested anonymity. The calling party is here identified as the
station originating the call. In order for this service to be
dependable, the called party must be able to trust that the calling
identity information being presented is valid. One way to achieve
this is described in [10].
To initiate a customer-originated-trace from an untrusted UAC, an
additional header is defined for the INVITE request. This header is
called P-DCS-Trace-Party-ID, and does not appear in any other request
or response. The entity addressed by the Request-URI performs the
service-provider-specific functions of recording and reporting the
caller identity in the P-DCS-Trace-Party-ID for law enforcement
action. It then forwards the call to either an announcement server
or to the service-provider's business office to collect further
information about the complaint. A trusted UAC does not use this
header, as it initiates this action locally.
5.1. Syntax
The ABNF description of this header is (some terms used in this ABNF
are defined in [2]):
P-DCS-Trace-Party-ID = "P-DCS-Trace-Party-ID" HCOLON
name-addr
This document adds the following entry to Table 2 of [2]:
Header field where proxy ACK BYE CAN INV OPT REG
------------ ----- ----- --- --- --- --- --- ---
P-DCS-Trace-Party-ID R dr - - - o - -
SUB NOT REF INF UPD PRA
--- --- --- --- --- ---
- - - - - -
The addr-spec contained in name-addr contains a URL that identifies
the remote endpoint. Addr-spec typically contains a tel: URL or SIP
URI giving the identity of the remote endpoint, as provided in the
signaling messages that established the session to be traced.
5.2. Procedures at an Untrusted User Agent Client (UAC)
The UAC MUST insert a P-DCS-Trace-Party-ID header into the initial
INVITE message for a customer-originated-trace request. The UAC MUST
use a SIP URI in the Request-URI with userinfo set to "call-trace"
and hostport identifying the call tracing entity for the untrusted
UA.
5.3. Procedures at a Trusted User Agent Client (UAC)
A trusted UAC performs the customer-originated-trace in a manner
similar to the trusted UAS, described below. A trusted UAC MUST NOT
include this header in any request.
5.4. Procedures at an Untrusted User Agent Server (UAS)
This header MUST NOT appear in any response sent by a UAS.
5.5. Procedures at a Trusted User Agent Server (UAS)
If the P-DCS-Trace-Party-ID header is present in the initial INVITE
request from a UAC, and the Request-URI of the INVITE has userinfo
set to "call-trace" and hostport set to the UAS, the UAS MUST perform
the service-provider-specific functions of recording and reporting
the caller identity for law enforcement action. The UAS then MUST
redirect the call, via a 3xx response, to either an announcement
server or to the service-provider's business office to collect
further information about the complaint.
This header MUST NOT appear in any response sent by a UAS.
5.6. Procedures at Proxy
Two sets of proxy procedures are defined: (1) the procedures at an
originating proxy, and (2) the procedures at a terminating proxy. The
originating proxy is a proxy that received the INVITE request from a
non-trusted endpoint.
The terminating proxy is a proxy that sends the INVITE request to a
non-trusted endpoint.
A proxy that both receives the INVITE request from an untrusted
endpoint, and sends the INVITE request to an untrusted endpoint,
performs both sets of procedures.
5.6.1. Procedures at Originating Proxy
If the P-DCS-Trace-Party-ID header is present in the initial INVITE
request from the UAC, and the Request-URI of the INVITE has userinfo
other than "call-trace" and hostport set to other than a potentially
provisioned call tracing entity, then the Proxy MAY reject the
request, or MAY remove the P-DCS-Trace-Party-ID header from the
request. If the header is present in a valid request, and contains a
private-URL that identifies the Proxy in the hostport, then the
Originating Proxy SHOULD replace the private-URL with its original
contents (i.e., the verified identity of the caller of the session
that is being traced).
5.6.2. Procedures at Terminating Proxy
This header MUST NOT appear in any request or response sent by a
terminating proxy to an untrusted endpoint.
6. P-DCS-OSPS
Some calls have special call processing requirements that may not be
satisfied by normal user agent call processing. For example, when a
user is engaged in a call and another call arrives, such a call might
be rejected with a busy indication. However, some PSTN operator
services require special call processing. In particular, the Busy
Line Verification (BLV) and Emergency Interrupt (EI) services
initiated by an operator from an Operator Services Position System
(OSPS) on the PSTN network have such a need. Similarly, emergency
calls to a 9-1-1 Public Service Access Point (PSAP) may result in
trunk signaling causing operator ringback using a howling tone or
sustained ring on the originating line (country-specific variations
may exist).
In order to inform the SIP user agent that special treatment should
be given to a call, we use a new P-DCS-OSPS header field, which may
be set to a value indicating when a special type of call processing
is requested. We define three values in this header, namely "BLV"
for busy line verification, "EI" for emergency interrupt, and "RING"
for operator ringback (e.g., howling/sustained tone ring in the US).
If the user agent decides to honor such a request, the response of
the user agent to an INVITE with either "BLV" or "EI" will not be a
busy indication. Since "EI" and "RING" only occur on established
dialogs, they may also appear in UPDATE requests.
6.1. Syntax
The ABNF description of the P-DCS-OSPS header is as follows (some
terms used in this ABNF are defined in [2]):
P-DCS-OSPS = "P-DCS-OSPS" HCOLON OSPS-Tag
OSPS-Tag = "BLV" / "EI" / "RING" / token
This document adds the following entry to Table 2 of [2]:
Header field where proxy ACK BYE CAN INV OPT REG
------------ ----- ----- --- --- --- --- --- ---
P-DCS-OSPS R dr - - - o - -
SUB NOT REF INF UPD PRA
--- --- --- --- --- ---
- - - - o -
The OSPS-Tag value of "token" is defined for extensibility, and is
reserved for future use.
6.2. Procedures at an Untrusted User Agent Client (UAC)
The P-DCS-OSPS header MUST NOT be sent in a request from an untrusted
UAC.
6.3. Procedures at a Trusted User Agent Client (UAC)
This header is typically only inserted by a Media Gateway Controller
[6] that is controlling a Media Gateway with special trunks to a PSTN
OSPS system or PSAP. This trunk group is usually referred to as a
BLV-trunk group and employs special signaling procedures that prevent
inadvertent use. Calls originating at the PSTN OSPS system are sent
over this trunk group, and result in an INVITE request with the P-
DCS-OSPS header.
This header MAY be sent in an INVITE request, and MUST NOT appear in
any message other than those listed below.
OSPS-Tag value "BLV" MUST NOT appear in any request or response other
than an initial INVITE request establishing a new dialog.
OSPS-Tag value "EI" MUST NOT appear in any request or response other
than (1) a subsequent INVITE within a pre-existing dialog established
with the OSPS-Tag value of "BLV", or (2) an UPDATE request within a
pre-existing dialog established with the OSPS-Tag value of "BLV".
OSPS-Tag value "RING" MUST NOT appear in any request or response
other than (1) a subsequent INVITE within a pre-existing dialog
established by a UAC to an operator or PSAP, or (2) an UPDATE request
within a pre-existing dialog established by a UAC to an operator or
PSAP.
6.4. Procedures at an Untrusted User Agent Server (UAS)
If the UAS receives an INVITE request with an OSPS-Tag of "BLV",
dialog identification that matches an existing dialog, and the
existing call was not established with the OSPS-Tag, it MUST reject
the request with a 403-Forbidden error code.
If the UAS receives an INVITE/UPDATE request with an OSPS-Tag value
of "EI" or "RING", with dialog identification that does not match an
existing dialog, it MUST reject the request with a 403-Forbidden
response code.
If the UAS receives an INVITE that contains an OSPS-Tag value of
"BLV" and is not willing to cooperate in offering this service, it
MUST reject the request with a 403-Forbidden response code.
The UAS SHOULD NOT reject an INVITE with a BLV OSPS-Tag due to a busy
condition. The UAS MUST NOT respond with a 3xx-Redirect response
code to an INVITE with a BLV OSPS-Tag. The UAS SHOULD NOT alert the
user of the incoming call attempt if the BLV OSPS-Tag is present in
the INVITE.
If an INVITE with OSPS-Tag of "BLV" is accepted (e.g., meeting all
QoS pre-conditions, etc.), the UAS MUST send an audio stream on this
connection to the address and port given in the SDP of the INVITE.
The UAS MAY perform a mixing operation between the two ends of an
existing active call and send the resulting media stream to the
address and port indicated. Alternatively, the UAS MAY send a copy
of the local voice stream, and (if no activity on the local voice
stream) send a copy of the received voice stream of an existing call.
If the state of the UAS is idle, the UAS SHOULD send a stream of
silence packets to OSPS. If the state of the UAS is ringing or
ringback, the UAS SHOULD send a ringback stream to OSPS.
If an INVITE/UPDATE with OSPS-Tag of "EI" is accepted, the UAS MUST
enable communication between the UAC and the local user. The UAS MAY
put any existing call on hold, or initiate an ad-hoc conference.
If an INVITE/UPDATE with OSPS-Tag of "RING" is accepted, the UAS MUST
perform operator ringback in accordance with local procedures, e.g.,
generate a 3-second howling tone or a sustained ring, depending on
the state of the user equipment.
6.5. Procedures at a Trusted User Agent Server (UAS)
The procedures at a trusted UAS MUST be identical to those described
in 6.4.
6.6. Procedures at Proxy
In the DCS architecture, the OSPS is considered a trusted UAC. If a
proxy receives a P-DCS-OSPS header in a request from an untrusted
source, it MUST either remove the header or reject the request with a
403-Forbidden response.
A proxy that implements a call-forwarding service MUST NOT respond to
an INVITE request with a 3xx response, if the request contained the
P-DCS-OSPS header.
7. P-DCS-BILLING-INFO
There are many billing models used in deriving revenue from telephony
services today. Charging for telephony services is tightly coupled
to the use of network resources. It is outside the scope of this
document to discuss the details of these numerous and varying
methods.
Proxies have access to subscriber information and act as policy
decision points and trusted intermediaries along the call signaling
path. Edge routers provide the network connection and resource
policy enforcement mechanism and also capture and report network
connection and resource usage information. Edge routers need to be
given billing information that can be logged with Record Keeping or
Billing servers. The proxy, as a central point of coordination
between call signaling and resource management, can provide this
information based on the authenticated identity of the calling and
called parties. Since there is a trust relationship among proxies,
they can be relied upon to exchange trusted billing information
pertaining to the parties involved in a call.
For Usage Accounting records, it is necessary to have an identifier
that can be associated with all the event records produced for the
call. The SIP Call-ID header field cannot be used as such an
identifier since it is selected by the originating user agent, and
may not be unique among all past calls as well as current calls.
Further, since this identifier is to be used by the service provider,
it should be chosen in a manner and in a format that meets the
service provider's needs.
Billing information may not necessarily be unique for each user
(consider the case of calls from an office all billed to the same
account). Billing information may not necessarily be identical for
all calls made by a single user (consider prepaid calls, credit card
calls, collect calls, etc). It is therefore necessary to carry
billing information separate from the calling and called party
identification. Furthermore, some billing models call for split-
charging where multiple entities are billed for portions of the call.
The addition of a SIP General Header Field allows for the capture of
billing information and billing identification for the duration of
the call.
It is the intent that the billing extensions would only appear on
trusted network segments, and MAY be inserted by a proxy or trusted
UA in INVITE requests in a trusted network segment, and removed
before leaving trusted network segments. The P-DCS-Billing-Info
header extension is used only on requests and responses between
proxies and trusted User Agents. It is never sent to, nor sent by,
an untrusted UA.
7.1. Syntax
The DCS-Billing-Info header is defined by the following ABNF (some
terms used in this ABNF are defined in [2]):
P-DCS-Billing-Info = "P-DCS-Billing-Info" HCOLON
Billing-Correlation-ID "/" FEID
*(SEMI Billing-Info-param)
Billing-Correlation-ID = 1*48(HEXDIG)
FEID = 1*16(HEXDIG) "@" host
Billing-Info-param = RKS-Group-ID-param / Charge-param /
Calling-param / Called-param /
Routing-param / Loc-Routing-param /
generic-param
RKS-Group-ID-param = "rksgroup" EQUAL RKS-Group-ID
RKS-Group-ID = token
Charge-param = "charge" EQUAL Acct-Charge-URI
Acct-Charge-URI = LDQUOT addr-spec RDQUOT
Calling-param = "calling" EQUAL Acct-Calling-URI
Acct-Calling-URI = LDQUOT addr-spec RDQUOT
Called-param = "called" EQUAL Acct-Called-URI
Acct-Called-URI = LDQUOT addr-spec RDQUOT
Routing-param = "routing" EQUAL Acct-Routing-URI
Acct-Routing-URI = LDQUOT addr-spec RDQUOT
Loc-Routing-param = "locroute" EQUAL Acct-Loc-Routing-URI
Acct-Loc-Routing-URI = LDQUOT addr-spec RDQUOT
This document adds the following entry to Table 2 of [2]:
Header field where proxy ACK BYE CAN INV OPT REG
------------ ----- ----- --- --- --- --- --- ---
P-DCS-Billing-Info admr - - - o - -
SUB NOT REF INF UPD PRA
--- --- --- --- --- ---
- - - - - -
The P-DCS-Billing-Info extension contains an identifier that can be
used by an event recorder to associate multiple usage records,
possibly from different sources, with a billable account. It further
contains the subscriber account information, and other information
necessary for accurate billing of the service. This header is only
used between proxies and trusted User Agents.
The Billing-Correlation-ID is specified in [9] as a 24-byte binary
structure, containing 4 bytes of NTP timestamp, 8 bytes of the unique
identifier of the network element that generated the ID, 8 bytes
giving the time zone, and 4 bytes of monotonically increasing
sequence number at that network element. This identifier is chosen
to be globally unique within the system for a window of several
months. This MUST be encoded in the P-DCS-Billing-Info header as a
hexadecimal string of up to 48 characters. Leading zeroes MAY be
suppressed.
The Financial-Entity-ID (FEID) is specified in [9] as an 8-byte
structure, containing the financial identifier for that domain,
followed by a domain name. FEID can be associated with a type of
service and could be assigned to multiple domains by the same
provider. A domain could contain multiple assigned FEIDs. This 8-
byte structure MUST be encoded in the P-DCS-Billing-Info header as a
hexadecimal string of up to 16 characters. Trailing zeroes MAY be
suppressed. "Host" contains the domain name.
The RKS-Group-ID specifies a record keeping server (or group of
cooperating servers) for event messages relating to this call. It is
used to control certain optimizations of procedures when multiple
event message streams are being sent to the same Record Keeping
Server.
Additional parameters contain the information needed for generation
of event message records. Acct-Charge-URI, Acct-Calling-URI, Acct-
Called-URI, Acct-Routing-URI, and Acct-Location-Routing-URI are each
defined as URLs; they should all contain tel: URLs with E.164
formatted addresses. These fields are further defined in [9] under
the element identifiers "Charge_Number" (element ID 16),
"Calling_Party_Number" (element ID 4), "Called_Party_Number" (element
ID 5), "Routing Number" (element ID 25), and
"Location_Routing_Number" (element ID 22).
7.2. Procedures at an Untrusted User Agent Client (UAC)
This header is never sent to an untrusted UAC, and is never sent by
an untrusted UAC.
7.3. Procedures at a Trusted User Agent Client (UAC)
The UAC MUST generate the Billing-Correlation-ID for the call, and
insert it into the P-DCS-Billing-Info header in the initial INVITE
message sent to the terminating proxy, along with the charging
information for the call. The UAC MUST include its FEID, and the
RKS-Group-ID for the Record-Keeping-Server being used by the UAC. If
the UAC performed a Local Number Portability (LNP) query, it MUST
include the Routing Number and Location Routing Number returned by
the query.
If the response to the initial INVITE is a 3xx-Redirect, the UAC
generates a new initial INVITE request to the destination specified
in the Contact: header, as per standard SIP. If a UAC receives a
3xx-Redirect response to an initial INVITE, the new INVITE generated
by the UAC MUST contain the P-DCS-Billing-Info header from the 3xx-
Redirect response. If the UAC is acting as a B2BUA, instead of
generating a new INVITE it MAY generate a private-URL and place it in
the Contact header of a 3xx-Redirect response sent to the originating
endpoint. This private-URL MUST contain (or contain a pointer to)
the P-DCS-Billing-Info value, which indicates the charging
arrangement for the new call, and an expiration time very shortly in
the future, to limit the ability of the originator to re-use this
private-URL for multiple calls.
A UAC that includes a Refer-to header in a REFER request MUST include
a P-DCS-Billing-Info header in the Refer-to's URL. This P-DCS-
Billing-Info header MUST include the accounting information of the
initiator of the REFER.
7.4. Procedures at an Untrusted User Agent Server (UAS)
This header is never sent to an untrusted UAS, and is never sent by
an untrusted UAS.
7.5. Procedures at a Trusted User Agent Server (UAS)
The UAS MUST include a P-DCS-Billing-Info header in the first
reliable 1xx (except 100) or 2xx response to an initial INVITE
message. This P-DCS-Billing-Info header MUST include the Billing-
Correlation-ID generated by the UAS, the FEID of the UAS, and the
RKS-Group-ID of the Record-Keeping-Server being used by the UAS. The
UAS MAY change the values of Acct-Charge-URI if it wishes to override
the billing information that was present in the INVITE (e.g., for a
toll-free call). The decision to do this and the contents of the new
Acct-Charge-URI MUST be determined by service provider policy
provisioned in the UAS. If the UAS performed a LNP query, it MUST
include the Routing Number and Location Routing Number returned by
the query.
The UAS MUST add a P-DCS-Billing-Info header to a 3xx-redirect
response to an initial INVITE, giving the accounting information for
the call forwarder, for the call segment from the destination to the
forwarded-to destination.
7.6. Procedures at Proxy
Three sets of proxy procedures are defined: (1) the procedures at an
originating proxy, (2) the procedures at a terminating proxy, and (3)
the procedures at a tandem proxy.
The originating proxy is a proxy that received the INVITE request
from a non-trusted endpoint.
The terminating proxy is a proxy that sends the INVITE request to a
non-trusted endpoint.
A proxy that is neither an originating proxy, nor a terminating
proxy, is a tandem proxy.
For purposes of mid-call changes, such as call transfers, the proxy
that receives the request from a non-trusted endpoint is considered
the initiating proxy; the proxy that sends the request to a non-
trusted endpoint is considered the recipient proxy. Procedures for
the initiating proxy are included below with those for originating
proxies, while procedures for the recipient proxy are included with
those for terminating proxies.
A proxy that both receives the INVITE request from an untrusted
endpoint, and sends the INVITE request to a non-trusted endpoint,
performs both sets of procedures.
7.6.1. Procedures at Originating Proxy
The originating proxy MUST generate the Billing-Correlation-ID for
the call, and insert it into the P-DCS-Billing-Info header in the
initial INVITE message sent to the terminating proxy, along with the
charging information for the call. The originating proxy MUST
include its FEID, and the RKS-Group-ID for the Record-Keeping-Server
being used by the originating proxy. If the originating proxy
performed a LNP query, it MUST include the Routing Number and
Location Routing Number returned by the query. Any P-DCS-Billing-
Info header present from an untrusted UA MUST be removed.
If the Request-URI contains a private-URL, and the decoded username
contains billing information, the originating proxy MUST generate a
P-DCS-Billing-Info header with that decrypted information. Otherwise,
the originating proxy MUST determine the accounting information for
the call originator, and insert a P-DCS-Billing-Info header including
that information.
If the response to the initial INVITE is a 3xx-Redirect, received
prior to a 18x, the originating proxy generates a new initial INVITE
request to the destination specified in the Contact: header, as per
standard SIP. If an originating proxy receives a 3xx-Redirect
response to an initial INVITE prior to a 18x response, the INVITE
generated by the proxy MUST contain the P-DCS-Billing-Info header
from the 3xx-Redirect response.
If the response to the initial INVITE is a 3xx-Redirect, received
after a 18x, the originating proxy generates a private-URL and places
it in the Contact header of a 3xx-Redirect response sent to the
originating endpoint. This private-URL MUST contain (or contain a
pointer to) the P-DCS-Billing-Info value, which indicate the charging
arrangement for the new call, and an expiration time very shortly in
the future, to limit the ability of the originator to re-use this
private-URL for multiple calls.
An originating proxy that processes a REFER request from an untrusted
UA MUST include a P-DCS-Billing-Info header in the Refer-to's URL.
This P-DCS-Billing-Info header MUST include the accounting
information of the initiator.
7.6.2. Procedures at Terminating Proxy
The terminating proxy MUST NOT send the P-DCS-Billing-Info header to
an untrusted destination.
The terminating proxy MUST include a P-DCS-Billing-Info header in the
first reliable 1xx (except 100) or 2xx response to an initial INVITE
message. This P-DCS-Billing-Info header MUST include the Billing-
Correlation-ID generated by the terminating proxy, the FEID of the
terminating proxy, and the RKS-Group-ID of the Record-Keeping-Server
being used by the terminating proxy. The terminating proxy MAY
change the values of Acct-Charge-URI if it wishes to override the
billing information that was present in the INVITE (e.g., for a
toll-free call). The decision to do this and the contents of the
resulting P-DCS-Billing-Info header MUST be determined by service
provider policy provisioned in the terminating proxy. If the
terminating proxy performed a LNP query, it MUST include the Routing
Number and Location Routing Number returned by the query.
The terminating proxy MUST add P-DCS-Billing-Info headers to a 3xx-
redirect response to an initial INVITE, giving the accounting
information for the call forwarder, for the call segment from the
destination to the forwarded-to destination.
A proxy receiving a mid-call REFER request that includes a Refer-to
header generates a private-URL and places it in the Refer-to header
sent to the endpoint. This private-URL MUST contain the P-DCS-
Billing-Info value, which indicate the charging arrangement for the
new call, and an expiration time very shortly in the future, to limit
the ability of the endpoint to re-use this private-URL for multiple
calls.
7.6.3. Procedures at Tandem Proxy
If the tandem proxy performed a LNP query, it MUST insert the Routing
Number and Location Routing Number returned by the query into the P-
DCS-Billing-Info header in the first reliable 1xx/2xx/3xx (except
100) response.
8. P-DCS-LAES and P-DCS-REDIRECT
NOTE: According to RFC 2804 [5], the IETF supports documentation of
lawful intercept technology if it is necessary to develop it. The
following section provides such documentation. The RFC 2119
language, as stated above, describes the requirements of the
specification only if implemented, and strictly within the
applicability domain described above. See RFC 2804 for description
of issues regarding privacy, security, and complexity in relation to
this technology.
The P-DCS-LAES extension contains the information needed to support
Lawfully Authorized Electronic Surveillance. This header contains
the address and port of an Electronic Surveillance Delivery Function
for delivery of a duplicate stream of event messages related to this
call. The header may also contain an additional address and port for
delivery of call content. Security key information is included to
enable pairs of Delivery Functions to securely exchange surveillance
information. This header is only used between proxies and trusted
User Agents.
The P-DCS-Redirect extension contains call identifying information
needed to support the requirements of Lawfully Authorized Electronic
Surveillance of redirected calls. This header is only used between
proxies and trusted User Agents.
Use of P-DCS-LAES and P-DCS-Redirect is controlled by a combination
of legislation, regulation, and court orders, which MUST be followed.
In certain cases inclusion of these headers will be mandated, and
therefore MUST be present in the requests and responses indicated.
In other cases inclusion of these headers will be forbidden, and
therefore MUST NOT be present in the request and responses indicated.
In the sub-sections that follow, use of "SHOULD" is intended to
capture these conflicting situations, e.g., a P-DCS-LAES header
SHOULD be included in an initial INVITE means either that it MUST be
included or that it MUST NOT be included, based on the applicable
court orders.
8.1. Syntax
The formats of the P-DCS-LAES and P-DCS-Redirect headers are given by
the following ABNF (some terms used in this ABNF are defined in [2]
and [3]):
P-DCS-LAES = "P-DCS-LAES" HCOLON Laes-sig
*(SEMI Laes-param)
Laes-sig = hostport
Laes-param = Laes-content / Laes-key / generic-param
Laes-content = "content" EQUAL hostport
Laes-key = "key" EQUAL token
P-DCS-Redirect = "P-DCS-Redirect" HCOLON Called-ID
*(redir-params)
Called-ID = LDQUOT addr-spec RDQUOT
redir-params = redir-uri-param / redir-count-param /
generic-param
redir-uri-param = "redirector-uri" EQUAL Redirector
Redirector = LDQUOT addr-spec RDQUOT
redir-count-param = "count" EQUAL Redir-count
Redir-count = 1*DIGIT
This document adds the following entry to Table 2 of [2]:
Header field where proxy ACK BYE CAN INV OPT REG
------------ ----- ----- --- --- --- --- --- ---
P-DCS-LAES adr - - - o - -
P-DCS-Redirect adr - - - o - -
SUB NOT REF INF UPD PRA
--- --- --- --- --- ---
- - - - - -
- - - - - -
The values of Laes-sig and Laes-content are addresses of the
Electronic Surveillance Delivery Function, and used as the
destination address for call-identifying information and call-
content, respectively. Laes-key is a string generated by the proxy
that is used by the Delivery Function to securely transfer
information between them [8].
The P-DCS-Redirect header contains redirection information. The
redir-uri-param indicates the original destination requested by the
user (e.g., dialed number), the Redirector indicates the new
destination, and the Redir-count indicates the number of redirections
that have occurred.
8.2. Procedures at an Untrusted User Agent Client (UAC)
This header MUST NOT be sent to an untrusted UAC, and MUST NOT be
sent by an untrusted UAC.
8.3. Procedures at a Trusted User Agent Client (UAC)
The UAC checks for an outstanding lawfully authorized surveillance
order for the originating subscriber, and, if present, includes this
information in the Authorization for Quality of Service [7] or
signals this information to the device performing the intercept
(e.g., a Media Gateway).
If the P-DCS-LAES header is present in the first reliable 1xx (except
100), 2xx or 3xx response (indicating surveillance is required on the
terminating subscriber, but that the terminating equipment is unable
to perform that function), the UAC MUST include this information in
the Authorization for Quality of Service, or MUST signal this
information to the device performing the intercept (e.g., a Media
Gateway).
If a 3xx-Redirect response is received to the initial INVITE request,
and if a P-DCS-LAES header is present in the 3xx response, the UAC
SHOULD include that header unchanged in the reissued INVITE. The UAC
SHOULD also include a P-DCS-Redirect header containing the original
dialed number, the new destination number, and the number of
redirections that have occurred. Although it is technically possible
for the originating equipment to perform this surveillance (or add to
its existing surveillance of the call), the design of the
surveillance system has the terminating equipment performing the
surveillance for all the intermediate forwardings.
A UAC that includes a Refer-to header in a REFER request, when the
originating subscriber has an outstanding lawfully authorized
surveillance order, SHOULD include a P-DCS-LAES header attached to
the Refer-to. The P-DCS-LAES header SHOULD include the address and
port of the local Electronic Surveillance Delivery Function for a
copy of the call's event messages, SHOULD include the address and
port of the local Electronic Surveillance Delivery Function for the
copy of call content if call content is to be intercepted, and SHOULD
include a random string for use as a security key between the
Delivery Functions.
The trusted UAC MUST NOT send the P-DCS-LAES and P-DCS-Redirect
headers to an untrusted entity.
8.4. Procedures at an Untrusted User Agent Server (UAS)
This header MUST NOT be sent to an untrusted UAS, and MUST NOT be
sent by an untrusted UAS.
8.5. Procedures at a Trusted User Agent Server (UAS)
The UAS checks for an outstanding lawfully authorized surveillance
order for the terminating subscriber, or presence of the P-DCS-LAES
header in the INVITE request. If either is present, the UAS includes
this information in the authorization for Quality of Service [7].
If the terminating equipment is unable to perform the required
surveillance (e.g., if the destination is a voicemail server), the
UAS SHOULD include a P-DCS-LAES header in the first reliable non-100
response requesting the originating proxy to perform the
surveillance. The P-DCS-LAES header SHOULD include the address and
port of the local Electronic Surveillance Delivery Function for a
copy of the call's event messages, SHOULD include the address and
port of the local Electronic Surveillance Delivery Function for the
copy of call content if call content is to be intercepted, and SHOULD
include a random string for use as a security key between the
Delivery Functions.
If the response to the initial INVITE request is a 3xx-Redirect
response, and there is an outstanding lawfully authorized
surveillance order for the terminating subscriber, the UAS SHOULD
include a P-DCS-LAES header in the 3xx-Redirect response, with
contents as described above.
The trusted UAS MUST NOT send the P-DCS-LAES and P-DCS-Redirect
headers to an untrusted entity.
8.6. Procedures at Proxy
Two sets of proxy procedures are defined: (1) the procedures at an
originating proxy, and (2) the procedures at a terminating proxy. The
originating proxy is a proxy that received the INVITE request from a
non-trusted endpoint.
The terminating proxy is a proxy that sends the INVITE request to a
non-trusted endpoint.
For purposes of mid-call changes, such as call transfers, the proxy
that receives the request from a non-trusted endpoint is considered
the initiating proxy; the proxy that sends the request to a non-
trusted endpoint is considered the recipient proxy. Procedures for
the initiating proxy are included below with those for originating
proxies, while procedures for the recipient proxy are included with
those for terminating proxies.
A proxy that both receives the INVITE request from an untrusted
endpoint, and sends the INVITE request to a non-trusted endpoint,
MUST NOT generate P-DCS-LAES nor P-DCS-Redirect headers.
A proxy that is neither an originating proxy nor a terminating proxy
SHOULD pass the P-DCS-Laes and P-DCS-Redirect headers in requests and
responses.
8.6.1. Procedures at Originating Proxy
The Originating Proxy MUST remove any P-DCS-LAES and P-DCS-Redirect
headers in requests or responses to or from an untrusted proxy or
untrusted UA.
The originating proxy checks for an outstanding lawfully authorized
surveillance order for the originating subscriber, and, if present,
includes this information in the Authorization for Quality of Service
[7] or signals this information to the device performing the
intercept (e.g., a Media Gateway).
If the P-DCS-LAES header is present in the first reliable 1xx (except
100), 2xx or 3xx response (indicating surveillance is required on the
terminating subscriber, but that the terminating equipment is unable
to perform that function), the originating proxy MUST include this
information in the Authorization for Quality of Service, or MUST
signal this information to the device performing the intercept (e.g.,
a Media Gateway).
If the Request-URI in an initial INVITE request contains a private-
URL, the originating proxy MUST decrypt the userinfo information to
find the real destination for the call, and other special processing
information. If electronic surveillance information is contained in
the decrypted userinfo, the originating proxy SHOULD generate a P-
DCS-LAES header with the surveillance information.
If a 3xx-Redirect response is received to the initial INVITE request
prior to a 18x, and if a P-DCS-LAES header is present in the 3xx
response, the originating proxy SHOULD include that header unchanged
in the reissued INVITE. The originating proxy SHOULD also include a
P-DCS-Redirect header containing the original dialed number, the new
destination number, and the number of redirections that have
occurred.
If a 3xx-Redirect response is received to the initial INVITE request
after a 18x, the originating proxy generates a private-URL and places
it in the Contact header of a 3xx-Redirect response sent to the
originating endpoint. If a P-DCS-LAES header is present in the 3xx
response, this private-URL MUST contain (1) the electronic
surveillance information from the 3xx-Redirect response, (2) the
original destination number, (3) the identity of the redirecting
party, and (4) the number of redirections of this call.
An originating proxy that processes a REFER request [4] from an
untrusted UA, when the originating subscriber has an outstanding
lawfully authorized surveillance order, becomes a B2BUA for that
request. It SHOULD reissue the request with a P-DCS-LAES header
added to the Refer-to's URL. The P-DCS-LAES header SHOULD include
(1) the address and port of the local Electronic Surveillance
Delivery Function for a copy of the call's event messages, (2) the
address and port of the local Electronic Surveillance Delivery
Function for the copy of call content if call content is to be
intercepted, and (3) a random string for use as a security key
between the Delivery Functions.
An initiating proxy that sends a mid-call REFER request including a
Refer-to header, when the initiating subscriber has an outstanding
lawfully authorized surveillance order, SHOULD include a P-DCS-LAES
header in the Refer-to's URL.
The originating proxy MUST NOT send the P-DCS-LAES and P-DCS-Redirect
headers to an untrusted entity.
8.6.2. Procedures at Terminating Proxy
The Terminating Proxy MUST remove any P-DCS-LAES and P-DCS-Redirect
headers in requests or responses to or from an untrusted proxy or UA.
The terminating proxy checks for an outstanding lawfully authorized
surveillance order for the terminating subscriber. If present, the
terminating proxy includes this information in the authorization for
Quality of Service [7].
The terminating proxy MUST NOT send the P-DCS-LAES and P-DCS-Redirect
headers to an untrusted entity, either as headers in the request or
response, or as headers attached to URIs in the request or response.
If the terminating equipment is unable to perform the required
surveillance (e.g., if the destination is a voicemail server), the
terminating proxy SHOULD include a P-DCS-LAES header in the first
reliable 1xx/2xx/3xx (except 100) response requesting the originating
proxy to perform the surveillance. The P-DCS-LAES header SHOULD
include the address and port of the local Electronic Surveillance
Delivery Function for a copy of the call's event messages, SHOULD
include the address and port of the local Electronic Surveillance
Delivery Function for the copy of call content if call content is to
be intercepted, and SHOULD include a random string for use as a
security key between the Delivery Functions.
If the response to the initial INVITE request is a 3xx-Redirect
response, and there is an outstanding lawfully authorized
surveillance order for the terminating subscriber, the terminating
proxy SHOULD include a P-DCS-LAES header in the 3xx-Redirect
response, with contents as described above.
A proxy receiving a mid-call REFER request [4] that includes a
Refer-to header with a P-DCS-LAES header attached becomes a B2BUA for
this request. It MUST generate a private-URL and place it in the
Refer-to header sent to the endpoint. This private-URL MUST contain
the P-DCS-LAES information from the attached header.
9. Security Considerations
QoS gate coordination, billing information, and electronic
surveillance information are all considered to be sensitive
information that MUST be protected from eavesdropping and furthermore
require integrity checking. It is therefore necessary that the
trusted UAs and proxies take precautions to protect this information
from eavesdropping and tampering. Use of IPsec or TLS between
Proxies is REQUIRED. A minimum mandatory-to-implement IPsec
configuration for the DCS architecture is given by [8]. Also
REQUIRED is mutual authentication (1) between Proxies and (2) between
trusted UAs and Proxies, both of which MAY be implemented with
administratively pre-shared keys, or through consultation with
another trusted third party. If IPsec is to be used, the
specification of the security policies and procedures of the
administrative domain where these headers are applicable (and all
connections between administrative domains in the federation) MUST
define an interoperable set of options.
10. IANA Considerations
This document defines a number of SIP extension headers, which have
been included in the registry of SIP headers defined in [2].
Registration information for new headers is as follows:
Header Field Name: P-DCS-Trace-Party-ID
RFC Number: 3603
Compact Form: none
Header Field Name: P-DCS-OSPS
RFC Number: 3603
Compact Form: none
Header Field Name: P-DCS-Billing-Info
RFC Number: 3603
Compact Form: none
Header Field Name: P-DCS-LAES
RFC Number: 3603
Compact Form: none
Header Field Name: P-DCS-Redirect
RFC Number: 3603
Compact Form: none
11. Intellectual Property Rights Notice
The IETF has been notified of intellectual property rights claimed in
regard to some or all of the specification contained in this
document. For more information consult the online list of claimed
rights.
12. References
12.1. Normative References
[1] Bradner, S., "Key words for use in RFCs to Indicate Requirement
Levels", BCP 14, RFC 2119, March 1997.
[2] 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.
[3] Crocker, D., Ed. and P. Overell, "Augmented BNF for Syntax
Specifications: ABNF", RFC 2234, November 1997.
[4] Sparks, R., "The Session Initiation Protocol (SIP) Refer
Method", RFC 3515, April 2003.
[5] IAB and IESG, "IETF Policy on Wiretapping", RFC 2804, May 2000.
12.2. Informative References
[6] DCS Group, "Architectural Considerations for Providing Carrier
Class Telephony Services Utilizing SIP-based Distributed Call
Control Mechanisms", Work in Progress.
[7] PacketCable Dynamic Quality of Service Specification, pkt-sp-
dqos-i07-030815, August 2003.
[8] PacketCable Security Specification, pkt-sp-sec-i09-030728, July
2003.
[9] PacketCable Event Message Specification, pkt-sp-em-i07-030815,
August 2003.
[10] Jennings, C., Peterson, J. and M. Watson, "Private Extensions to
the Session Initiation Protocol (SIP) for Asserted Identity
within Trusted Networks", RFC 3325, November 2002.
13. Acknowledgements
The Distributed Call Signaling work in the PacketCable project is the
work of a large number of people, representing many different
companies. The authors would like to recognize and thank the
following for their assistance: John Wheeler, Motorola; David
Boardman, Daniel Paul, Arris Interactive; Bill Blum, Jon Fellows, Jay
Strater, Jeff Ollis, Clive Holborow, Motorola; Doug Newlin, Guido
Schuster, Ikhlaq Sidhu, 3Com; Jiri Matousek, Bay Networks; Farzi
Khazai, Nortel; John Chapman, Bill Guckel, Michael Ramalho, Cisco;
Chuck Kalmanek, Doug Nortz, John Lawser, James Cheng, Tung- Hai
Hsiao, Partho Mishra, AT&T; Telcordia Technologies; and Lucent Cable
Communications.
Previous versions further acknowledged, as co-authors, several people
for providing the text of this document. They are:
Bill Marshall (wtm@research.att.com) and K. K. Ramakrishnan
(kkrama@research.att.com), AT&T; Ed Miller
(edward.miller@terayon.com), Terayon; Glenn Russell
(G.Russell@Cablelabs.com), CableLabs; Burcak Beser
(burcak@juniper.net) Juniper Networks, Mike Mannette
(Michael_Mannette@3com.com) and Kurt Steinbrenner
(Kurt_Steinbrenner@3com.com), 3Com; Dave Oran (oran@cisco.com) and
Flemming Andreasen (fandreas@cisco.com), Cisco Systems; John
Pickens (jpickens@com21.com), Com21; Poornima Lalwaney
(poornima.lalwaney@nokia.com), Nokia; Jon Fellows
(jfellows@coppermountain.com), Copper Mountain Networks; Doc Evans
(n7dr@arrisi.com) Arris, and Keith Kelly (keith@netspeak.com),
NetSpeak.
14. Editors' Addresses
Bill Marshall
AT&T
Florham Park, NJ 07932
EMail: wtm@research.att.com
Flemming Andreasen
Cisco
Edison, NJ
EMail: fandreas@cisco.com
15. Full Copyright Statement
Copyright (C) The Internet Society (2003). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
English.
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assignees.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Acknowledgement
Funding for the RFC Editor function is currently provided by the
Internet Society.