Note: Descriptions are shown in the official language in which they were submitted.
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
ROUTING MESSAGES
FIET~D OF THE INVENTION
The present invention relates to the routing of messages, and
in particular but not exclusively in an IMS system.
BACKGROUND TO THE INVENTION
The introduction of Third Generation (3G) communication
systems will significantly increase the possibilities for
accessing services on the Internet via mobile user equipment
(UE) as well as other types of UE.
Various user equipment (UE) such as computers (fixed or
portable), mobile telephones, personal data assistants or
organisers and so on are known o the ski-Iced person and can
be used to access the Internet to obtain services. Mobile
user equipment referred to as a mobile station (MS) can be
defined as a means that is capable of communication via a
wireless interface with another device such as a base station
of a mobile telecommunication network or any other station.
Such a mobile user equipment can be adapted for voice, text
message, data communication or multimedia communication via
the wireless interface.
The term "service" used above and hereinafter will be
understood to broadly cover any service or goods. which a user
may desire, require or be provided with. The term also will
be understood to cover the provision of complimentary
services. In particular, but not_ exclusively, the term
"service" will be, understood to include Internet multimedia
services, conferencing, telephony, gaming, rich call,
presence, e-commerce and messaging e.g. instant messaging.
1
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
The 3G Partnership Project (3GPP) is defining a reference
architecture for the Universal Mobile Telecommunication System.
(UMTS) core network which will provide the users of. UE with
access to these services. This UMTS core network is divided
into three principal domains. These are the Circuit Switched
domain, the Packet Switched domain and the Internet Protocol
Multimedia (IM) domain.
The latter of these, the IM domain, makes sure that multimedia
services are adequately managed. The IM domain supports the
Session Initiation Protocol (SIP) as developed by the Internet
Engineering Task Force (IETF).
SIP is an application layer signalling protocol for starting,
changing and ending user sessions as well as for sending and
receiving transactions. A session may, for example, be a two-
way telephone call or multi-way conference session or
connection between a user and an application server (AS). The
establishment of these sessions enables a user to be provided
with the above-mentioned services. One of the basic features
of SIP is that the protocol .enables personal mobility of a
user using mobile UE by providing the capability to:reach a
called party (which can be an application server AS) via a
single location independent address.
In this document the following abbreviations will be used:
AS Application Server
BGCF Breakout Gateway Control Function
CN Core Network
CPS Connection Processing Server
CS Circuit Switched
CSCF Call Session Control Function or Call State Control
Function
2
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
DNS Domain Name System
ENUM See "E.164 number and DNS" (RFC2916)
FQDN Fully Qualified Domain Name
GW/S/AS network function or entity e.g. a proxy, and/or Gateway
and/or Server and/or Application Server or the like
HSS Home Subscriber Server
I-CSCF Interrogating CSCF
ID Identity
IM IP Multimedia
IMS IP Multimedia core network Subsystem
IMS-WV-GW Gateway between IMS and WV networks
IP Internet Protocol
ISC IP multimedia Service Control
MGCF Media Gateway Control Function
NAPTR Naming Authority Pointer (RFC 2915)
O-CSCF Outbound CSCF
P-CSCF Proxy CSCF
PMG Presence (P) , Messaging (M) and Group Management (G)
PLS Presence List Server
PS Presence Server
PMG-WV-GW Gateway between IMS and WV networks
RR Resource Record of DNS
S-CSCF Serving CSCF
SIP Session Initiation Protocol (RFC 3261)
SIP URI SIP Uniform Resource Identifier (RFC 3261)
SLF Subscription Locator Function
SSR Service and Subscription Repository
TEL URL Is an URL associated to a terminal that can be
contacted using the telephone network (RFC 2806)
UE User Equipment
UMS User Mobility Server
UMTS Universal Mobile Telecommunications System
URI Uniform Resource Identifier
URL Uniform Resource Locator
WV Wireless Village ,
3
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Terminating sessions/transactions are routed in an IMS from
the I-CSCF to an S-CSCF that can route them to an AS following
the rules of a ffilter criteria. If the target identity (i.a.
public user identity). is not registered, the I-CSCF selects an
S-CSCF, and the S-CSCF down loads filter criteria from the
HSS. However there is a problem where the target identity is
not an IMS identity - non-IMS identities are routed over the
IMS network to a non-IMS network.
An AS originated session/transaction is routed in IMS from AS
to an S-CSCF that can route them further. Normally this S-
CSCF is the one that was used when the session/transaction was
routed from S-CSCF to AS, or address of the S-CSCF that is
returned from the HSS or other database as response to a
query, or address of the (default) S-CSCF may be configured in
AS or fetched from an internal or external database, table,
list, configuration data storage or alike. There are cases
where it is difficult or impossible to find an S-CSCF.
Here are some examples where it can be difficult to find a S-
CSCF:
a) If the subscriber is not registered, possibly no S-CSCF is
assigned to the subscriber (or more accurately to any public
user identity of the subscriber).
b) If the sending network element is a service server that
routes a session/transaction on behalf of the user, there is a
similar situation i . a . there may be no S-CSCF assigned to the
user. (This kind of service server is referred to as. user
dependent service server).
c) If ,a third party user uses a group identity as target
address e.g.. a message is sent to a group by a user that is
not the _"owner" of the group identity, there is a problem in
4
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
deciding which S-CSCF should be used when the group server
sends a message to each member of the group.
d) If the sender is a service that has no connection to any
user (i.e. the sender is a user independent service server).
At least in this case the AS has to choose an S-CSCF or use a
default S-CSCF. Both solutions have drawbacks. In the first
one the AS has to perform functionalities of I-CSCF i.e.
choose an S-CSCF. In the second one (i.e. if the default is
1
used), the problem is how the load is balanced (Round robin
functionality in DNS may be used.)
An additional argument against routing through an S-CSCF is
that no service of S-CSCF is needed e.g. no filter criteria is
utilized. This is especially true in the user independent
service server case.
Routing with service identities is another problem of IMS. In
order to route to an AS, server, gateway, network function,
network entity or alike that hosts or offers the service, an
entry is needed in SLF and HSS containing routing information
(e. g. filter criteria) for routing to S-CSCF and from S-CSCF
to the correct AS, server, gateway, network function, network
entity or the like that hosts or offers the service. The
result is that HSS has to contain. all service identities with
proper routing information. There is a similar problem with
group identities created by users. A user may for example
create a group of work colleagues, a group of family and a
group of friends. These identities with proper routing
information have to be included in HSS. Service identities may
be quite stable but the group identities may be changed
relatively often. A group identity may be a list of users that
can be used e.g. to send a message to all of them with a
single message sending procedure (instead of repeating the
procedure in order to send the same message to every one of
5
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
them). The problem of using a service and group identity is
the creation/modification/deletion of a more or less temporary
entry in HSS in order to make the routing possible via an S
CSCF to a proper AS, server, gateway, network function,
network entity or alike.
It has also been found that when a Presence List Server (PLS)
subscribes to the presence information of presentities, the
routing done according to the current 3GPP IMS standard is not
optimal. In addition, when the PLS (AS) initiates a request by
itself, it is not defined how the PLS (AS) selects an S-CSCF.
There exists a problem that if a group server is seen as an
application server, an ISC interface should be used. This has
the disadvantage that routing is complicated in that an S-CSCF
is needed in both the terminating and originating cases.
Another problem is that in known arrangements, the application
server has to store all used service identities into an SLF,
an HSS and/or another subscriber database.
SUMMARY OF THE INVENTION
It is an aim of embodiments of the present invention to
address one or more of the problems described.
According to one aspect of the invention, there is provided a
method of routing for a message via an IMS system comprising
the steps of receiving a message at an I-CSCF, obtaining
address information for a network function for which said
message is intended and sending said message to said network
function accordance with said address information.
6
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
The network function may be provided by a network entity. The
network function may be an application server, server, gateway
or any other suitable entity.
Preferably, said message is sent directly or via a proxy or
gateway element to the network function via a gateway element.
Preferably, said obtaining step comprises querying a database.
Preferably, said database comprises a SLF.
Preferably, said database provides said address information
for said network function.
Preferably, said database provides information identifying a
further database.
Preferably, said further database may comprise one of a HSS,
UMS or SSR.
Preferably, said further database contains said address
information.
Preferably, said further database contains configuration
information of said network function.
Preferably, the method comprises the step of determining if
said message is for an IMS target or a non-IMS target.
Preferably, said steps are followed only if it is determined
that said message is not routed to any S-CSCF because the said
message is for a non IMS target, or for an IMS target. that
identifies a service or a group or the like.
7
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
According to one aspect of the invention, there i.s provided. a
method of routing a message from a network function via an IMS
system comprising the steps of: : ..
originating a message from a network function ;
determining the address of a proxy entity to which said
message is to be sent;
routing said message to said proxy entity; and ~ .
routing said message from said proxy entity to an entry
point of a target network.
Preferably, said entry point is in the same 'or different
network as said network function.
Preferably, wherein said originating step . comprises
originating one of a session and a transaction.
Preferably, said determining step comprises the step of
querying one of a database, table, file and a list.
Preferably, said determining step comprises determining the
proxy entity from information contained in said network
function. ' ..
Preferably the method comprises the step of determining the
entry point to which said message is to be routed.w .
Preferably, said proxy entity is arranged to determ~.ne the
entry point to which said message is to be sent.
Preferably, said proxy entity is arranged to"determine the I-
CSCF to which said message is to be sent by ~accessing'a
database.
Preferably, said database comprises a DNS.
8
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
According to one aspect of -the invention, there is provided a
method of 'routing a message from a network function,via_an IMS
system comprising the step's of originating a message from a
network 'function determining' the I-CSCF to which said message
is to be sent, routing said message .directly to said I=,CSCF if.
said I-CSCF is in a same network as said network function.
According to one aspect of the invention, there is provided a
method of routing a message from a network functiom via an IMS
system comprising the steps of originating a message from a
network function, determining the I-CSCF to which said message
is to be sent, routing said message directly to said, I-CSCF if
said I-CSCF is in a trusted network
According to one aspect of the invention, there is provided a
method of routing a message from a network function via an IMS
system, said method comprising the steps of sending a request
from the network function to an I-CSCF, determining at the I-
CSCF - the S-CSCF to which a message from said network function
is to be sent and sending said. message to the determined S-
CSCF.
Preferably, said network function comprises a PLS.
Preferably, said determining step comprises querying a
database.
Preferably, said determining step comprises querying.a HSS.
According to one aspect of the invention, there is provided a
method of routing a message from a first network function via
an IMS system, said method comprising the steps of-sending a
request from the first network function to an I-CSCF,
9
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
determining at the I-CSCF a second network function to which a
message from said first network function is to be sent and
sending said message directly'from the I-CSCF to said second
network function..
According to one aspect of the invention, there is provided a
method of routing a message comprising the steps of receiving
a message in accordance with a first protocol, converting said
message to a second protocol, querying a database using
identification information in said message to obtain new
identification information and using said new identification
information to route the message to a proxy.
Preferably, said proxy is arranged to route said message.
Preferably, said proxy is arranged to obtain a translation of
said identity.
Preferably, said proxy routes the message to another network.
Preferably, the proxy routes the message to an I-CSCF.
Preferably, an I-CSCF is arranged to query said database.
Preferably, said I-CSCF is arranged to route said message to
said proxy.
Preferably, an entity receiving said message is arranged to
route said message to said proxy..
Preferably, wherein said second'protocol is SIP.
Preferably,' said proxy is arranged to route said message to a
gateway.
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
BRIEF DESCRIPTION OF FIGURES
For a better understanding of the present invention and as to
how the same may be carried into effect, reference will be
made by way of example only to the accompanying drawings in
which:
Figure 1 shows a known method of normal terminating routing in
an IMS system;
Figure 2a shows a method embodying the present invention of
routing in an IMS system;
Figure 2b shows schematically routing with non-IMS schemes;
Figure 2c shows schematically routing from WV user equipment;
Figure 2d shows routing from WV domain to. either WV or IMS
domain;
Figure 3a shows a known method of routing in an IMS system,
where the session or transaction originates with the AS;
Figure 3b shows a method embodying the present invention in an
IMS system, where the session or transaction originates with
the AS;
Figure 4 shows a signal flow of a further embodiment of the
invention;
Figure 5a shows a first arrangement where routing is done with
an O-CSCF;
11
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Figure 5b shows a second arrangement where routing is done
with an O-CSCF;
Figure 5c shows a third arrangement where routing is done with
an 0-CSCF;
Figure 5d shows a fourth arrangement where routing is done
with an O-CSCF;
Figure 6a shows a known arrangement for routing where a group
server is an application server and there is a subscriber
initiated group session;
Figure 6b shows a known arrangement where a group server is an
application server and there is a group server initiated group
session;
Figure 6c shows a embodiment of the present invention where
the group server is not an application server and there is a
subscriber initiated group session;
Figure 6d shows an example embodying the present invention
where the group server is not an application server and the
group server initiates a group session;
Figure 7a shows an arrangement where a server offers
subscriber independent services, in the originating case;
Figure 7b illustrates where the server offers subscriber
independent services in the terminating case:
Figure 8a to c show three routing scenarios for requests that
originate from a public service identity PSI, embodying the
invention; and
12
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Figure 9 shows the flow of messages where the PSI is the
originator, embodying the invention. .
. DETAILED DESCRIPTION OF EMBODIMENTS OF THE INVENTION
Embodiments of the present invention will be described in
relation to a UMTS system in accordance with the so-called
third generation standards. Reference is made to the following
third generation partnership project standards which are
hereby incorporated by reference. These documents describe the
IP multimedia system to which embodiments of the present
invention are particularly applicable. However the embodiments
of the present invention are also applicable to any other type
of SIP network, regardless of whether or not it is an IMS
network as well as to non SIP networks which may or may not be
IMS networks.
3GPP TS 23.002: "Network architecture".
3GPP TS 23.228: "IP multimedia subsystem; Stage 2".
3GPP TS 24.229: "IP Multimedia Call Control Protocol based on
SIP and SDP; Stage3"
3GPP 23.841 Presence Service; Architecture and Functional
Description
3GPP 24.841' Presence based on SIP; Functional models, flows
and protocol details
Embodiments of the present invention may use SIP'. In order to
provide access to the Internet and other IM services to users,
protocols have been developed to assist in providing telephony
and multimedia services across the Internet. The session
initiation protocol (SIP) is one such protocol which has been
developed for controlling the creation, modification and
termination of sessions with one or more parties. The call
13
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
sessions may include e.g. Internet or other IP network
telephone calls, conferences or other multimedia services and
activities.' The transactions may include e.g. Internet or
other IP network messaging, presence, group, and . other
multimedia services and activities.
SIP addressing follows the popular Internet convention of
identifying a user by a unique address using Uniform Resource
Locators (URL's) or as defined in RFC3261 it is SIP URI. SIP
signalling between two users consists of a series of, requests
and responses. A SIP transaction has dual parties,-the user
agent client (UAC) who sends a request and a user agent server
(UAS) who responds in reply to the request. The client and
server comprise the SIP user agent. In addition to this, SIP
includes the SIP network server which is the network devices
which handle signalling associated with multiple calls.
As is known in the art an SIP invitation typically includes
two messages. It will be understood that there may be more
messages than only these and that, in fact, in 3GPP there are
more messages used. These are not discussed herein for the
sake of brevity. The two messages are an INVITE, initiated by
the caller UAC and a 200 OK message from the callee. This
latter message is typically acknowledged by the caller after
which stage the parties may communicate according to
parameters sent and received during signalling. Both the
caller and callee can end a session by executing a BYE
message. During an established session a new; set of
parameters may be selected by either participant producing a
further INVITE message or by using some other SIP message.
In the following, references are made to application servers.
In alternative embodiments, the application server may instead
14
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
be a network function or entity e.g. a proxy or a gateway or a
server or the like.
Embodiments of the invention aim to avoid finding a S-CSCF for
sessions/transactions that do not need any actions or services
offered by the S-CSCFs but only routing to/from a GW/S/AS i..e.
a network function or entity e.g. a proxy and/or~gateway
and/or server and/or application server or the like. To do
this routing is done directly from I-CSCF to a proper GW/S/AS
with the help of SLF and/or HSS that returns address of the
proper GW/S/AS. Embodiments of the invention can be applied at
least to the following cases:
a) routing of non-IMS schemes via IMS network to a non-IMS
network (e.g. WV scheme)
b) routing of generic schemes to a correct network: to IMS or
to non-IMS (e. g. PRES- presence and IM instant message
schemes)
c) routing of service and group identities to a proper GW/S/AS
Generic schemes are protocol'independent schemes specifying
only the service but not the protocol. For example "IM"
specifies the service to be "Instant Messaging" but not the
used protocol that would e.g. be SIP in case of IMS.
Respectively "pres" specifies the "Presence" service.
If non-IMS identities are not inserted into the SLF and/or
HSS, it is possible to use for example pseudo entries.in the
SLF and/or HSS to handle the normal routing via the IMS
network to an AS (i.e. from I-CSCF to S-CSCF (filter criteria)
to AS). The filter criteria is needed in the S-CSCF to choose
the correct AS to which to route the non-IMS identity. An
example of routing non-IMS identities via an IMS to non-IMS
network are Wireless Village (WV) identities that are routed
from I-CSCF to an S-CSCF and further to an AS or server that
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
acts as gateway e.g. .IMS-WV-GW (or IMS Gateway) to WV network.
These are discussed in more detail hereinafter.
Reference will now be made to Figure 1 which shows how routing
is currently carried out in known IMS systems with the signal
flow indicated diagrammatically. At least some of the messages
may be in accordance with the SIP (session initiated
protocol). These messages are shown in capitals.
An I-CSCF 100 receives a message in step S1 such as an initial
INVITE or MESSAGE.
The I-CSCF 100 then sends a query to the SLF 102. in step S2
and the SLF returns the address of the correct HSS 104. If
there is only one HSS, SLF is not needed, and step S2 can be
omitted.
In step S3, the I-CSCF 100 then sends a query to the
identified HSS 104. The HSS 104 replies with the address of
the correct S-CSCF 108 or the capabilities of a needed S-CSCF.
If needed, the I-CSCF selects an S-CSCF.
In step S4, the I-CSCF 100 routes the message to the S-CSCF
108. The S-CSCF down loads routing information, (e. g. filter
criteria for routing to application servers) if not yet
downloaded..
In step S5, the S-CSCF 108 routes the message to the correct
application server 106 using the downloaded routing
information.
Reference will now be made to Figure 2a, which. shows the
routing used in a first embodiment of the invention. In
16
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
particular, the routing of non-IMS schemes and service/group
identities and dynamic identities is shown. It should be
appreciated that in Figure 2a, the routing of terminating
sessions/transaction from the I-CSCF to the GW/S/AS are shown.
The same reference numbers will be used for the same entities
as shown in Figure 1.
In step T1, the I-CSCF 100 receives a message such as initial
INVITE or MESSAGE.
In step T2, the I-CSCF 102 makes a query to the SLF/HSS 102.
The SLF/HSS returns the address of the correct GS/S/AS 106.
Optionally, the SLF/HSS 102 may return the address of the
database or server such as a HSS, UMS user mobility server or
SSR subscriber service router or repository or database
containing dynamic public user identities or database
containing dynamic service identities or other database.
SLF/HSS denotes SLF, or HSS in the case there is no SLF.
In step T3, the I-CSCF 102 will optionally make a query to the
database 110 identified in step T2. It should be noted that
the SLF may have returned the actual address of the database
or its configuration information or the like. The database
will return the address of the correct GW/S/AS 106.
In step T4, the I-CSCF 100 routes the message to the correct
GW/S/AS using the address returned by the SLF/HSS, if
provided, or if not from the database 110.
Non-IMS schemes are other schemes than those associated with
the user, group or service identities of IMS i:e. currently
"sip" and "tel", which are also the originally used schemes in
IMS. "Non-IMS scheme" is used in embodiments of the invention
to refer to schemes which are not currently IMS schemes. As
17
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
the standard evolve, it is of course possible that a current
so-called non-IMS scheme may become an IMS scheme. If a non-
IMS scheme becomes an IMS scheme, embodiments of the invention
may still apply.'
In embodiments of the invention, the following may be done:
1. Routing to the target IMS network with non-IMS
scheme is done normally only if the target operator
allows it to be done.
2. Routing via the target IMS network to WV network is
done by routing directly from I-CSCF to PMG-WV-GW or
any other IMS gateway because non-IMS subscriber has
- normally no entry in HSS
- no filter criteria
nothing to do with IMS
IMS is thus only a routing path to WV network
3. No fallback if faulty scheme
Normally error is returned to UE
IMS does not normally correct the faulty scheme
4. wv:+3584022334455Cdomain,
im:+3584022334455Cdomain,
press+3584022334455~domain,
sip:+3584022334455@domain
are valid WV routable identities.
wv:+3584022334455 is valid WV identity in the domain
in question
When non-IMS scheme is present, it is normally checked whether
the target operator will receive the message via SIP. To do
this, a . g. a DNS query is made with target domain name .. It is
asked for SRV records e.g. with _im.-sip.operator.net. The
answer might be e.g. _im.-sip.operator.net SRV 0. 0 5060 i-
CSCf.operator.net.
18
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
This answer indicates that "im" scheme is accepted with SIP
protocol in the port 5060 of the address "i-
cscf.operator.net".
The target operator may or may not allow messages using a
non-IMS scheme. If the target operator allows the non-IMS
scheme, a routable address is found with DNS query and the
message will be routed to that address. The scheme is not
normally modified.
If the target operator will not receive the non-IMS scheme,
that is no routable address is found with DNS query, the
message will be routed to the appropriate GW/S/AS e.g. PMG-WV-
GW or IMS gateway. The filter criteria are not used to find
the correct GW/S/AS and the address of GW/S/AS is configured
in S-CSCF or is fetched from a table, list or database or the
like. This can be done using for example a routing table as
follows:
schema target
wv pmg-wv-gw.home.net
pres pmg-wv-gw.home.net
im pmg-wv-gw.home.net
19
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
When a non-IMS scheme is present, it is checked whether the
message should be routed to a S-CSCF e.g. because the target
identity is an IMS identity. The I-CSCF makes a SLF and/or HSS
query. The scheme may or may not be carried over Cx and Dx
interfaces (standardized). There may be different ranges for
different schemes or the schemes may be marked somehow inside
the subscriber data.
If the message should be routed to a S-CSCF for example the
identity is found to be an IMS identity in the SLF/HSS, the
schema is handled according to the general principles of IMS
and routing is as normal terminating case in IMS.
If the message should not be routed to a S-CSCF for example
the identity is not found to be an IMS identity in SLF/HSS,
the I-CSCF finds the correct GW/S/AS where the message is
routed. The GW/S/AS address is returned from SLF/HSS or the
address of GW/S/AS is configured in I-CSCF. No S-CSCF is
involved. There may be a new interface between I-CSCF and
GW/S/AS e.g. PMG-WV-GW or other IMS gateway.
Routing to WV is possible via target IMS domain.
Reference is made to Figures 2b to 2d which illustrate the
above-described scenario. Referring first to Figure 2b:
This is where there is an IMS ID. From the user equipment 500
a message is sent in step D1 to the P-CSCF 502 which in turn
sends a message in step D2 to the S-CSCF 506. Next, the S
CSCF makes a query with the DNS 504 in step D3. In response
to that query, the S-CSCF sends a message in step D6 to the I
.,,
CSCF 514. The I-CSCF 514 sends message in step D7 to the SLF
508 and receives a reply. In the next step, the I-CSCF sends
a message to the HSS and receives a reply in step D8. In step
D9, the I-CSCF 514 sends a message to S-CSCF 516. The steps'
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
D7 to D9 are as .described earlier as steps S2 to S4 in
relation to Figure 1.
Where there is a non-IMS ID, the following steps occur: in
particular, steps D1, D2, D3, D6, D7 and optionally D8 are as
described when there is an IMS identity. The next step
however is step D11 where the I-CSCF 514 contacts the second.
PMG-WV-GW or IMS gateway 520. The second IMS gateway 520
sends a message to a second WV server 526 in step D12. In
step D14, the WV server 526 sends a message to WV user
equipment 528.
Where routing is not possible via the target IMS, the route ..
taken is the same as steps D1 to D3 already described.
However, the next step is then D4 where the S-CSCF 506 sends a
message to the first IMS gateway 522. The next step may
either be step D5 or DSb. In step D5, a message is sent to a
first WV server 524 which contacts the second WV server 526 in
step D13. A message is sent in step D5b directly from the IMS
gateway 522 to the second WV server 526. In both cases the
next step will be step D14 where the second WV server sends a
message to the WV user equipment 528.
It should be appreciated that the gateway entities 522, 520
and PMG 518 can all be regarded as proxies or servers, and for
example application servers.
Reference is made to figure 2C which shows where the routing
from the WV server is configurable (based on the scheme and/or
domain). Normally, routing via the WV is the first choice and
routing to the IMS is the second choice. However, this may be
reversed in some embodiments of the present invention.
21
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Routing to the WV may be via the target IMS domain. This is a
result of the DNS query at the outbound proxy O-CSCF. Routing
after the outbound proxy is the same as described in relation
to figure 2b. Those elements, which are the same .in figure
2b, are marked with the same reference numbers.
Where there is a routable URI, that is routing is possible via
the target IMS, there are two options. The first one is where
there is an IMS ID. In this case, routing from the first WV
user equipment.528a is then to the first WV server 524 in step
E1. The first WV server 524 sends a message in step E2 to the
first IMS gateway 522 which in turns sends a message in step
E3 to the O-CSCF 530. The O-CSCF 530 sends a message in step
E4 to the DNS 504. In response to the information received
from DNS 504, the O-CSCF 530 sends a message in step E6 to the
I-CSCF 516. The I-CSCF 516 obtains information from the SLF
508 in step E7 and information from the HSS 510 in step E8. E8
may also be an alternative to step E7 if there is no SLF.
Next, the I-CSCF contacts in step E9 the S-CSCF 516 identified
by steps E7 and/or E8.
Where there is a non IMS ID, steps El, E2, E3, E4, E6 and E7
are performed as described in relation to the IMS ID. Step E8
is optional and/or an alternative to step E7 if there is no
SLF. The next step is then step E11 where the I-CSCF 516
contacts second IMS gateway 520. In step E12, the second IMS
gateway 520 contacts the second WV server 526. In step E14,
the second WV server 526 contacts the second WV user equipment
528 b. .
Where routing is not possible via the target IMS,._the. steps
taken would be steps E1, E2, E3 and E4. At that point, an
error would be returned to WV server 524 that may for example
try to route to the target WV server 526.
22
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Again, the first and second IMS gateways 522 and 520 as well
as PMG server 518 may be proxies or servers, for example
application servers.
Reference is made to Figure 2d which illustrates routing'to
WV/IMS via the WV domain. Again, the same reference numerals
are used for the same entities.. At the target WV server 526,
it is checked whether the user is a WV user. Where the user
is a non WV user the following steps occur. Firstly, in step
F1, a message is sent from the source WV user equipment 528b
to the first WV server 524. The first WV server 524 sends a
message in step F13 to the second WV server 526. The second
WV server 526 sends a message in step F12 to the second IMS
gateway 520 which in turn sends a message to the I-CSCF 514 in
step F11. The I-CSCF obtains information from the SLF 508 in
step F7 and information from HSS 510 in step F8 as previously
described.
Next, in step F9, the I-CSCF 514 sends a message in step F9 to
the S-CSCF 516. The S-CSCF 516 sends a message in step F10 to
the user equipment 512 or the PMG server 518.
Where a WV user 528a is the target user, a much simpler
routing occurs.. The WV user equipment, which is the .source,
528b, sends a message in step F1 to the first WV server 524
which then sends a message in step F13 to the second WV server
526. The second WV server 526 in step F14 sends a message to
the WV user equipment which is the target, that is WV UE 528a.
Loop detection is needed in the I-CSCF or in the IMS gateway
or WV server routes to IMS gateway only messages with target
identities of IMS or the IMS gateway changes the scheme to SIP
to prevent further routing back to the WV network.
23
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
An example will now be given of information stored in the SLF
in one embodiment of the invention:
An entry in SLF may contain e.g. the following information:
a) Usual IMS entry to refer to the proper HSS:
john. smith@operator.net reference to HSS_3
b) John Smith has also WV subscription. WV traffic is routed
to WV network through the gateway IMS-WV-GW:
wv:john.smithCoperator.net reference to IMS-WV-GW
c) John Smith wants to receive Instant messages in WV.network:
im:john.smith@operator.net reference to IMS-WV-GW
d) John Smith wants to offer his Presence information from IMS
network:
pres:john.smith~operator.net reference to HSS 3
e) John Smith has created a group (consisting of his fishing
friends) to be used e.g. with message services. For example
anyone can send an Instant message to the whole group by
sending it to the group identity:
fishingfriends..john.smithCoperator.net reference to
group server
The entry to offer movie services also to customers from other
networks might contain the following information: .
movies@operator.net reference to movie-server
These are only examples of information. References to a HSS
and to a gateway or server must differ in order that the I-
CSCF is able to act accordingly: to make HSS query to the
24
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
indicated HSS or to route the message to the indicated gateway
or server respectively.
Embodiments of the invention avoid allocating an S-CSCF to
route non-IMS identities over the IMS to non-IMS networks.
Advantages of the embodiments of the invention are:
a) No big influence on HSS because no filter criteria is
needed.
b) No need to allocate an S-CSCF i.e. saves resources.
c) No 'influence on S-CSCF i.e. no need to have special '
scenarios for handling non-IMS identities.
d) SLF can contain all non-IMS identities or as on option a
pseudo entry only for one or more groups of the non-IMS
identities.
e) Operator can offer service to its IMS customers to
prioritize IMS or non-IMS services e.g. presence service is
offered from the WV network (when the IMS subscriber also has
WV subscription).
f) Group names and service names can be routed directly to the
correct GW/S/AS. They need no entry in HSS.
g) Routing to several GW/S/ASs is difficult. To solve the
problems discussed above, routing to one GW/S/AS is enough. Of
course the SLF may return several addresses if needed. These
could be tried one after another until a GW/S/AS is found that
accepts the message. These addresses could also be used as a
route through several GW/S/ASs.
Scheme can be "visible" in SLF and/or in HSS i.a. it is part
of the key that is used to identify entries in SLF and/or in
HSS or every public user identity entry in SLF and/or in HSS
indicates what are the valid schemes with that public user
identity.
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
A second embodiment of the. present invention will now be
described with reference to Figures 3a and 3b.
. In order to avoid finding a S-CSCF for a session/transaction,
where no S-CSCF is allocated or it is difficult to find out
the address of the allocated S-CSCF, the session/transaction
is routed from the GW/S/AS to an O-CSCF i.e. outbound proxy.
From O-CSCF i.e. outbound proxy the session/transaction is
routed further to I-CSCF of the target network.
The O-CSCF i.e. the outbound proxy may be bypassed when the
target network is the same network i.e. the target I-CSCF is
located in the same network or in a trusted network. In this
case session/transaction is routed directly from GW/S/AS to
the I-CSCF.
This embodiment of invention can be applied at least to the
following cases:
a) routing from GW/S/AS (e.g. from IMS WV gateway) non-IMS
traffic (e.g. with WV scheme) via IMS network to another IMS
network
b) routing from GW/S/AS sessions/transactions where the
originator (e.g. service group server) is loosely or not at
all connected to any subscriber; (in this case GW/S/AS is
referred as user independent service server)
GW/S/AS may start the session/transaction or GW/S/AS may be a
gateway passing traffic to IMS network. Both cases are
referred here as GW/S/AS originated.sessions/transactions.
In embodiments of the invention, the address or the name of
the GW/S/AS, O-CSCF (i.e. outbound proxy) and I-CSCF may be
configured in GW/S/AS. Addresses or names may also be, fetched
from a database (e. g. DNS), table, file, server or the like.
26
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
The addresses and names may be resolved e.g. with database
(e. g. DNS), table, file, server or the like.
In general in all embodiments the addresses and/or names of
the functions, gateways, servers, elements and the other
entities of a network may be resolved e.g. with database (e. g.
DNS), table, file, server or the like.
The 0-CSCF is a logical functionality that may be implemented
with I-CSCF in the same network element. Alternatively, the
functionality of the I-CSCF may be changed so that it includes
the functionality of the O-CSCF.
This embodiment is concerned with avoiding finding and/or
allocating an S-CSCF to route GW/S/AS originated
sessions/transactions.
Reference will now be made to Figure 3a which shows known
routing in an IMS system.
In step A1, the GW/S/AS 204 originates a session or
transaction. The session or transaction is routed to a S-CSCF
202. The address of the S-CSCF may 'be known from the previous
session or transaction or it may be queried from the HSS or it
may be configured in the GW/S/AS.. Possibly the filter criteria
are evaluated and the session or transaction may be routed to
an AS according to the filter criteria.
The next step is either step A2a or A2b. In step A2b, the
session or transaction is routed to an I-CSCF 200 in the same
network as the S-CSCF. In step A2a, the session or transaction
is routed to an I-CSCF 206 in a different network to the S-
CSCF 202.
27
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Reference is now made to Figure 3b, which illustrates a second
embodiment of the invention. In particular, the routing of
non-IMS identities and routing of cases with IMS service/group,
identity as an originator is shown. It should be appreciated
that in Figure 3b, the routing of originating
sessions/transaction from the GW/~S/AS to the 0-CSCF (i,.e.
outbound proxy) are shown.
0-CSCF is an outbound proxy that may be like S-CSCF without
subscriber database, because the 0-CSCF normally does not need
to perform any activities associated to IMS subscribers. O-
CSCF may have all other features of the S-CSCF.
In step Bla, the AS 204 originates a session or transaction.
The session or transaction is routed to O-CSCF 208. The
address of the O-CSCF is queried from a database or the like
or is fetched from a table, a file, a list or the like or is
configured in the GW/S/AS.
Step Blb is an optional step and allows optimal routing from
the GW/S/AS 204 directly to the I-CSCF 200 if the I-CSCF is
located in the same network or in a trusted network.
The next step is either step B2a or B2b. In step B2b, the O-
CSCF routes the session or transaction to a I-CSCF 200 in the
same network whilst in step B2a, the O-CSCF 208 routes the
session or transaction to an I-CSCF 206 in another network.
Advantages of this second embodiment of the invention are:
a) No influence on HSS, because there is no need to insert
service/group identities ,(and possibly also filter criteria)
to SLF and/or HSS in order to be able to allocate an S-CSCF
when a GW/S/AS originates a session/transaction on behalf of a
service/group identity.
28
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
b) No'need to allocate an S-CSCF i.e. saves resources.
c) No influence oi~. S-CSCF. No scenarios needed to align
Subscriber Profile Database (SPD) to handle these cases.
d) Sessions/transactions on behalf of service/group identities
5. can be routed directly from GW/S/AS to 0-CSCF without passing
any S-CSCF.
e) The solution can be seen as a second (GW/S/AS to I-CSCF in
the own network) and a third (GW/S/AS to O-CSCF) possibility
to route GW/S/AS originated sessions/transactions. The first
possibility is to route via S-CSCF if the S-CSCF can easily be
used.
When the Presence List Server (PLS) terminates some request
and it triggers a new request or some request is initiated by
the PLS, in the current 3GPP IMS architecture the PLS needs to
send the request to an S-CSCF. This can be done based on the
Record-Route header of the received request (if there was one)
or the PLS can have the S-CSCF configured. In a better
solution, the PLS can directly send the message to an I-CSCF
and leave out the S-CSCF as in case of PLS (group) there are
no originating services.
29
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Public service URIs (i.e. URIs that are identities of services
or alike) and group URIs (i.e.. URIs that are identities of
groups or alike) can be routed according to the embodiments of
this invention. In the first~embodiment (i.e. in .case where
routing from I-CSCF to GW/S/AS) the SLF/HSS may return name or
address of the GW/S/AS similarly as it does in cases where
routing with an non-IMS identity according to the embodiment.
In the second embodiment (i.e. GW/S/AS originated case) the
messages with a service URI or a group URI as an originator
are routed to O-CSCF similarly as messages with non-IMS
identity as an originator according to the embodiment.
Reference will now be made to ffigure 4 to describe a further
embodiment of the invention. Current 3GPP architecture
requires unnecessary involving of the S-CSCF where the S-CSCF
selection is problematic or not optimal.
The advantage of embodiments of the invention is that this
solution works in all cases and is simpler than the one that
follows the current 3GPP PMG architecture.
Figure 4 shows the messages in embodiments of the invention.
This can be summarised as follows:
Watcher agent in a UE subscribes to a presence list (SUBSCRIBE
#1 from UE-till PLS).
The request is answered (200 OK from PLS to UE) .
PLS initiates a subscription to one of the presentities
belonging to the list (SUBSCRIBE #2 PLS till PS).
This can be sent through the S-CSCF or as proposed in
embodiment of this invention, it can be directly sent to the
I-CSCF
The answer to the subscription is routed back from PS to PLS
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Reference is made to Figure 4 which shows the signal flow in
the third embodiment of the present invention. In step C1, a
subscribe message is sent from the user equipment 400 to the
P-CSCF 402. In step C2 the subscribe' message is sent from the
P-CSCF 402 to the S-CSCF 404. In step C3 the subscribed
message is sent from the S-CSCF 404 to the PLS (AS) 406. The
PLS(AS) sends in step C4 a 200 .OK message (which.is a SIP
acknowledgement message) back to the S-CSCF 404. In step C5,
the S-CSCF 404 sends the 200 OK message to the P-CSCF 402.
The P-CSCF 402 sends the 200 OK message in step C6 to the user
equipment 400. The flow shown in figure 4 now shows two
options.
The optimal signal flow will now be described as followed.
The next step is for a prescribed message to be sent in step
C7 from the PLS (AS) 406 to the I-CSCF 408. In step C8, there
is a HSS query where the I-CSCF 408 sends a query to the HSS
410 and receives a reply. The HSS will return the correct S-
CSCF or the capabilities of a needed S-CSCF.
In the next step C9 a second subscribe message is sent from
the I-CSCF 408 to the S-CSCF 412. The S-CSCF will send the
subscribe message to the PS 414 in step C10. The presence
server 414 sends an acknowledgement message such as a 200 OK
message in step C11 to the S-CSCF 412. The S-CSCF 412 .sends
in step C12 a 200 OK message to the I-CSCF 408. Finally, the
I-CSCF 40,8 sends a message in step C13 to the PLS (AS) in step
C13. This message is the 200 OK message.
In a less optimal solution, step C7 is replaced by steps C7a
and C7b. In those steps, the PLS (AS) 406 sends the second
subscribe message first to the S-CSCF 404 in step C7a. In
step C7b, the S-CSCF 404 sends the subscribe message to the I-
CSCF 408. Additionally, step C13 is replaced by steps Cl3a
31
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
and step Cl3b in this less optimal solution. In this less
optimal solution, the I-CSCF 408 sends the 200 OK message in
step Cl3a to the S-CSCF 404. In step Cl3b, the S-CSCF 404
sends the 200 OK -message to the PLS (AS). This second
solution is less optimal in that it is not clear to which S-
CSCF to send the messages if the PLS generates the request by
itself. .
The implementation is simple, as the PLS (defined in 3GPP to
be an AS) need to send the PLS originated requests to the I
CSCF instead of the S-CSCF.
Reference is now made to Figures 5a to 5d which show
arrangements in which the outbound proxy is utilised to get a
more optimal routing in the case of number~portability. The
outbound proxy has the capability to act as a CSCF without
subscriber profile database i.e. it does not handle
subscribers and thus no filter criteria connected to any
subscriber are down loaded from subscriber database e.g. HSS.
The outbound proxy, which is here called outbound CSCF i.e. O
CSCF, is capable of making the ENUM translation. It can also
route to similarly as a S-CSCF. .The O-CSCF can be used to
solve number portability routing problem with proposed
solution where MGCF should route directly to another network;
and I-CSCF routes directly to an I-CSCF in another network.
In embodiments of the invention, the MGCF~ routes the
session/transaction (with the new identity returned from
number portability database e.g. SLF) to an O-CSCF. The O-~CSCF
checks the identity to see whether ENUM translation is needed.
If it is, the O-CSCF performs the' translation. In ,short the O-
CSCF does all the same procedures as S-CSCF when it routes
session/transaction to another network. The main difference
between S-CSCF and 0-CSCF is the usage. S-CSCF is used when
32
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
there exists a user in the same network who can be linked to
the session/transaction; while O-CSCF is used when such a user
cannot be found. In the CS originated case the calling party
is not a subscriber of the IMS network. Because the target
number is ported to another network neither is the called
party is a subscriber of the network. Thus with using O-CSCF
it is possible to avoid routing through S-CSCF, and also to
avoid adding new functionalities to MGCF in order to make it
capable of routing directly to another network.
.
One modification is to let the I-CSCF route the
session/transaction directly to 0-CSCF instead of MGCF routing
to O-CSCF. The decision procedure in I-CSCF is simple: because
the new identity (after the number portability procedure) is
not identity of this network, the session/transaction has to
be routed out of this network to the target network. Thus
routing to an O-CSCF is an evident choice. No S-CSCF can be
naturally chosen, because the new identity is not linked to
any identity that could be registered in this network.
The same solution can be applied also to cases where number
portability procedure is done in the terminating network and
the session/transaction is received from another IMS network.
In Figures 5a and 5b, porting to the IMS domain is
illustrated.
The MGCF 500 receives a message from CS that is a call set up
message in step G1. The MGCF 500 converts the message to a SIP
message to for example and initial INVITE message.
In step G2, the MGCF 500 sends the message to the I-CSCF 502
in the same network which receives that message.
33
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
In step G3, the I-CSCF 502 makes a query to a~ number
portability database such as SLF 506 with the target number of
the initial INVITE message. _ .
In step G4, the SLF 504 returns the new identity.
In step G5, the I-CSCF 502 returns e.g. the "301 moved
permanently" response to the MGCF 500.
In step G6, the MGCF reroutes the message to the 0-CSCF 504.
In step G7, the O-CSCF 504 makes an ENUM translation of the
new identity if it is or contains a number, for example E.164.
This translation involves an ENUM entity 508.
In .step G8, the 0-CSCF 504 routes the message to a new I-CSCF
510 in another IMS network.
Reference will now be made to Figure 5b, which shows a
modification to the arrangement of Figure 5a. Those elements
which are the same as shown in Figure 5a are marked with the
same reference numbers.
Steps H1 to H4 correspond to steps G1 to G4 of Figure 5a.
In step H5, the I-CSCF 502 routes the message to the O-CSCF
504.
Steps H6 and H7 correspond to steps G7 and G8 respectively.
Reference is now made to Figures 5c and 5d which illustrate
porting to a CS domain.
Steps J1 to J7 correspond to steps G1 to G7 respectively.
34
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
In step J8, because the O-CSCF 504 is not able to get a
routable SIP URI, the O-CSCF 504 routes the message to a first
BGCF 514.
The next step is then J9a or J9b which constitute normal
routing steps. In particular, routing is either to a second
BGCF 516 in step J9b or to a second MGCF 512 in step J9a.
Figure 5d shows a modification to the arrangement of Figure
5c.
Steps K1 to K6 are the same as steps H1 to H6 respectively of
Figure 5b and steps K7, K8a and K8b correspond to steps J8,
J9a and J9b respectively.
In embodiments of the invention, the SLF/HSS may be queried by
the I-CSCF during a registration or session set-up to get the
name of the HSS containing the required subscriber specific
data or get the name of the an adaptation function such as an
application server, gateway or the like for further routing.
The notation SLF/HSS means SLF, or HSS if SLF does not exist.
The adaptation functionality is arranged to communicate with
the CSCF and performs protocol conversion between appropriate
protocols and the IM subsystem control protocols, if required.
The adaptation functionality is arranged to act as a gateway
or server where requests with non-SIP schemes may be routed.
The SLF/HSS may be arranged to handle the schemes and return a
SIP routable address as the name of the adaptation
functionality. ..
In embodiments of the present invention, the I-CSCF can send a
query e.g. DX SLF QUERY or' alike to the SLF/HSS and includes
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
as a parameter the URI which is stated in the INVITE request .
The SLF/HSS looks up in its database for the queried URI. The
SLF answers with the HSS name in which the user's subscription
data can be found or alternatively the SLF/HSS may answer with
the name of the adaptation functionality where the request
shall be routed.
The unknown status of a requested party can be determined in
the SLF/HSS. The I-CSCF requests information on the user to be
reached, identified by the URI included in the INVITE request
and the SLF/HSS responds back to the I-CSCF with an indication
that the user is unknown if it can not find the queried URI.
The I-CSCF uses the indication that the user is unknown
returned from the SLF/HSS to formulate the correct SIP message
back towards the originating party to indicate that the user
is unknown.
Communications between the CSCF and adaptation functionality
may be in accordance with the SIP protocol. A single session
control protocol may be applied to the interface between the
CSCF and the adaptation functionality. This may be the SIP
protocol or another protocol.
In embodiments of .the invention, the routing of SIP signalling
within the IMS may use URIs. The CSCFs and adaptation
functionality may be identifiable using a valid SIP URL (host
domain name or network address) on those interfaces supporting
the SIP protocol. These SIP URLs may be used when identifying
these nodes in header fields of messages.
Optionally SLF/HSS may return GW/S/AS address with. a new
identity. Thus SLF/HSS can be used as a portability network
entity or device that returns a new identity with routing
address.
36
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
A service URI is in the first place connected to a service. In
the second~place it may also be connected to an user (which
has caused its creation) for example because of Charging. A
service URI maybe used as an identity of the originator when
originating a session/transaction on behalf of a service. A
user may create a group identity:
Type I (subscriber initiated group session or transaction)
In this type of group session and transaction normally
everybody pays his own session to group session or
transaction. The procedure is that the user reserves a group
identity from the group server. The user sends the group
identity to members of the group and then the members initiate
the session or transaction to the group identity.
Type II (group server initiated group session or transaction)
In this type of group session and transaction normally the
originator pays all sessions to group session or transaction.
The procedure is that the user reserves a group identity from
the group server. The user sends a list of members to the
group server. The group server initiates the sessions or
transactions to the group members.
This will now be described in more detail with reference to
Figure 6. Figures 6a and 6b describe known routing
arrangements where the group server is an application server.
These figures are included to aid the explanation of
embodiments of the present invention. Figures 6c and 6d
illustrate embodiments of the present invention. In Figure 6c,
the routing form the I-CSCF to the group server is as in
Figure 2a. The routing from the group server to the I -CSCF or,
to the 0-CSCF is like the routing in Figure 3b.
37
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Reference is made to figure 6a which shows a Type I case i . a .
the subscriber initiated group session or transaction. The
originator 600 sends in step 1 a message to the P-CSCF 602
which in step 2 contacts a corresponding S-CSCF 604. The S-
CSCF contacts in step 3 the group server 606. In step 4 the
group server contacts the subscription database 608 which can
take any suitable form and may for example be an SLF, an HSS
or a DRR(Dynamic Resource Register) or a database capable to
store dynamic identities or alike. This effectively allows
the originator 600 to reserve a group identity. The group
identity is stored or activated in the subscriber database 608
by the group server 606 in step 4 and 5. The group server
returns the group identity to the' originator 600 via the S-
CSCF 604 and the P-CSCF 602 in steps 6, 7 and 8 respectively.
The originator 600 connects to the network which contains to
the group server 606.
The originator 600 then sends the group identity to the other
members of the group, that is entity B, referenced 618 and
entity C referenced 624. This is not shown. The originator
600 and entity B 618 both connect to the network which
contains the group server 606. Entity C 624 connects to a
different network to that containing the group server 606.
The members of the group, that is entity B 618 and entity C
624 initiate a session on the basis of the~group identity. In
step 21, each of the members 618 and 624 contact a P-CSCF 616
and 622 respectively. The respective P-CSCF contacts in step
22 a respective S-CSCF 614 and 620. The respective S-CSCF
contacts in step 23 a common I-CSCF 612. It should be
appreciated that the P-CSCF and S-CSCF associated with entity
B are~in the same network as the group server while the P-CSCF
and S-CSCF associated with entity C are in a different network
to the group server.
38
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
The I-CSCF 612 interrogates the subscriber database 608 to
obtain the relevant S-CSCF, this taking place in steps 24 and
25. The I-CSCF 612 then contacts the identified S-CSCF 610 in
step 26. That S-CSCF then contacts the group server 606. In
this way, the session is initiated.
Reference is now made to Figure 6b . which shows a type II case
i.e. the group server initiated group session or .transaction
and the group server is an application server. The elements
which are the same as shown in figure 6a are marked by the
same reference numerals. The originator 600 reserves a group
identity in steps 1 to 8, these steps being the same or
similar to those described in relation to Figure 6a. The
originator then sends a list of members to the group server.
This is not shown. The group server then initiates sessions
to the members in steps 21 to 29 which will now be described.
Again in this example the members are entity B 618 and entity
C 624 with entity B being connected to the same network which
contains the group server and entity C being connected to a
different network.
Firstly, the group server 600 contacts in step 21 to a S-CSCF
610. The S-CSCF 610 contacts in step 22 to the subscriber
database 608 to get the needed subscriber information e.g. the
originating filter criteria associated to the group identity.
This information is returned in step 23 to the S-CSCF 610.
The S-CSCF 610 contacts the appropriate I-CSCF 612 and 626.
The I-CSCF 612 for the entity B user interrogates the
subscriber database in step 25 and receives information on the
S-CSCF 614 to be used for the group member B, in step 26.
Likewise the I-CSCF 626 for entity C 624, contacts a HSS 628
in step 25 which provided information in step 26 on the S-CSCF
to be used.
39
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
The I-CSCFs 612 and 626 then contact the respective S-CSCF 614
and S-CSCF 620 for the respective members B and C 618 and 624.
This happens in step 27. In step 28, the respective S-CSCFs
614 and 620 contact respective P-CSCFs 616 and 622 associated
with the respective members B and C 618 and 624. The
respective P-CSCFs 616 and 622 then contact in step 29 the
members B and C 618 and 624 respectively. In this way, the
group server is able to initiate a session.
Reference is now made to figure 6c which shows a type I case
i.e.the subscriber initiated group session or transaction where
the group server is not an application server. Instead, the
group server may be a server. Again, those elements which are
the same as shown in figure 6a are marked with the same
reference numbers. In this arrangement, the group server is
referenced by 606'. Instead of a subscriber database, there
is a routing database 608'. The difference between 608 and
608' is the same as between 102 and 104 of figure 1 (normal
subscriber DB) and 102 and 110 of figure 2a. The routing
database can be provided by ari SLF and/or HSS and/or DRR.
In steps 1 to 8, the originator 600 reserves a group identity.
These steps are the same or similar to those described- in
relation to figures 6a and 6b. However, it should be
appreciated that steps 4 and 5 may be omitted in embodiments
of the present invention. In this case the group server 606'
may have the necessary group identity and not need to look it
up from the routing database or store it into the routing
database.
The originator 600 then sends the group identity to the
members of the group. Again, this is not shown. Next, the
members of the group 618 and 624 initiate a session to the
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
group identity in steps 21 to 26. Steps 21 to 25 are as
described in figure 6a except routing database 60;8' is used
instead of subscriber database 608. In practice there may be
little difference between a routing database and a subscriber
database. However, a connection is made directly from the I-
CSCF 612 to the group server 606' in step 26. This may be as
discussed in relation to previous embodiments.
Reference is made to figure 6d which shows an example where
the group server is not an application server and it is a type
II case, i.e. the group server initiated group session or
transaction . Again, those elements which are the same as in
figure 6a, b, and c are referred to with the same reference '
numbers.
The originator 600 reserves a group identity in steps 1 to 8.
Again, as with figure 6c, steps 4~and 5 may be omitted. The
originator then sends a List of members to the group server
606'. These steps are not shown for clarity.
The group server then initiates sessions with the members.
Steps 21 to 26 are an example where it is carried out with the
identity of a member in the group and the identity is SIP URI.
This allows members attached to the same network as the group
server to be contacted to establish a session. This is
illustrated by steps 21 to 26. In this, the group server 606'
connects directly to the I-CSCF 612, and not via an S-CSCF.
This is as discussed in relation to earlier embodiments. The
I-CSCF 612 interrogates the routing database 608 in step 22 to
receive routing information from the database in step 23. That
routing information may be the S-CSCF to which the message
from the I-CSCF 612 is to be routed. Based on that routing
information, the I-CSCF 612 contacts the S-CSCF 614 associated
with the user 618. This occurs in step 24. In step 25 the S-
41
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
CSCF 614 contacts the P-CSCF 616 associated with the entity B
618. In step 26, the P-CSCF 616 contacts the entity B 618.
In this way, a~session is initiated.
The group server may alternatively or additionally initiate
sessions to the member with a TEL URL of the own network.
This allows members attached to the same network as the group
server to be contacted to establish a session. In this step,
the group server 606' contacts in step 31 an 0-CSCF 630. In
step 32, the O-CSCF 630 looks up the ENUM from a database 632.
This occurs in step 32 with the reply being sent to the O-CSCF
in step 33. In step 34, the O-CSCF 630 contacts the I-CSCF
612. Steps 22 to 26 already described are then carried out.
In this way, the session can be established.
The group server can initiate a session alternatively or
additionally with a foreign TEL URL, that is with a TEL URL of
a different network. This allows members attached to a
different network as the group server to be contacted to
establish a session. In this, steps 31 to 34, already
described are carried out. However this case, step 34 would
allow the O-CSCF 630 to contact the I-CSCF 626 associated with
the member C 624. That I-CSCF 626 and member C are part of and
connected to respectively a different network to that
containing the group server. In step 35 the I-CSCF obtains
routing information for the subscriber 624 from the HSS 628.
The information is returned in step 36. This information may
identify the S-vCSCF 620 to be used. The I-CSCF 626 then
contacts the identified S-CSCF 620 in step 37. In step 38 the
S-CSCF 620 contacts the respective P-CSCF 622. The P-CSCF 622
contacts member C in step 39. In this way, a session is
established.
42
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Alternatively or additionally, the group server can initiate a
session with a member using a foreign SIP URI. In other
words, the SIP URI of a different network is used. This allows
members attached to a different network to the group server to
be contacted to establish a session. This involves step 31
and then steps 34 to 39 already described. In other words,
steps 32 and 33 are omitted.
In embodiments of the present invention the group server is
not an application server so an ISC interface may not be used.
In known arrangements, an ISC interface is bound .with a S-CSCF
and AS, that is there is an ISC interface between these
entities. To route to an AS involves going via an ISC, that is
S-CSCF to ISC to AS and vice versa. A filter criteria is used
at an S-CSCF to select and AS. In this embodiment, the aim of
the group server is to avoid the restriction that the routing
to an AS must always go from the S-CSCF via the ISC interface.
The point of the group server arrangement described in
relation to figure 6c and d and Figures 7a and b, is:
a) Routing to group server is allowed from S-CSCF optionally
with ISC and also via other interfaces in addition to the
optional ISC (e. g. via normal SIP).
b) Routing to group server is allowed also from other elements
in addition to the optional routing from S-CSCF in terminating
cases.
c) Routing from group server is allowed also to other elements
in addition to the optional routing to S-CSCF in originating
cases.
43
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
The group server is seen as an entry point to another network
(which can be regarded as the network offering group
sessions).
The group server can be regarded as being an I-CSCF of another
network. It is possible in some embodiments of the present
invention, that a group server can consist of application
server and non application server parts. Reserving group
entity is routed as to an application server whilst routing to
the group identity is as routing to a server. Both routings of
figure 1 and 2a are valid to the same server e.g. application
server. This has the advantage that routing becomes simple.
No S-CSCF involvement is required in cases where the group
server is originator of the session. No. HSS involvement is
necessary. The SLF can offer addresses to the group server in
the terminating case. SLF may contain wildcard entries that
are associated to the routing to a certain group server or
servers. The group servers) gives) out i.e. deliver(s) only
group addresses that match one of the wildcard entries in the
SLF. This way group identities need no to be stored as dynamic
identities to subscriber database (e. g. HSS, DRR or. alike).
As an example *.john.doe@operator.net may be a wildcard entry
in SLF (or in HSS if there is no SLF) . When John Doe: want to
reserve a group identity, the group server gives to John Doe
only group identities containing his own identity e.g.
fishing-friends.john.doe~operator.net and
family.john.doeCoperator.net. ~ .
Reference is now made to figure 7a which shows a server 606"
offering subscriber independent services. Figure 7a shows the
originating ease where routing is from the subscriber
independent server. Those elements which were the~same as
shown in figure 6 are referred to as the same reference
44
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
number. The same step number is used for those steps which
correspond to those shown in figure 6..
In the case where the users own SIP URI is used, steps 31, 34
and 22 to 26 already described in relation to .Figure 6 are
carried out in that order. In this case, instead of the group
server or application server shown in figure 6, there is a
subscriber independent server 606". As in figures 6a and b,
there is a subscriber database 608. In step 34, the O-CSCF
contacts the I-CSCF 612 in the same network as contains the
server.
In some embodiments of the present invention, this can be
optimised and steps 21 to 26 can be carried out when the own
SIP URI is used, leaving out steps 31 and 34.
If the network's own TEL URL is used, then steps 31 to 34 and
steps 22 to 26 are carried.out in that order. In step 34, the
O-CSCF contacts the I-CSCF 612 in the same network as contains
the server.
If the ENUM translation fails, routing can still take place
with the TEL URL . In this case, steps 31 to 34 and 41 to 42
are used. In step 34, a BGCF 650 is contacted by the O-CSCF
630. The BGCF 650 contacts an MGCF 652 in step 41 which in
turn connects in step 42 to the circuit switched domain 654.
Routing can be done with a TEL URL of a foreign or different
network. This involves steps 31 to 33 and 34 to 39 as
described previously. In steps 32 ENUM query is used to get
information in step 33 to resolve the TEL URL into SIP URI to
be used for routing. In step 34, the 0-CSCF connects to the I-
CSCF 626 of the different network to that containing the
server 606". -
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
If routing is carried out with a SIP URI from a different
network, then steps 31 and 34 to 39 are carried out, in that
order. In step 34, the O-CSCF connects to the I-CSCF 626 of
the different network to that containing the server 606".
It should be appreciated that in some embodiments, one or more
of these routing methods can be carried out. For example,
routing to different users can be carried out on this basis.
Reference is now made to figure 7b which shows the terminating
case. Again, those elements which are the same as shown in
figure 6 and 7a are referred to by the same reference number.
Where routing to the subscriber independent server 606" is.
from the same network as the server, the following steps are
carried out in this order. In step 121, the user 618 contacts
its associated P-CSCF 616 which in turn contacts the
appropriate S-CSCF 614 in step 122. In step 123, the~S-CSCF
614 contacts the I-CSCF 612. In step 124, the I-CSCF 612
contacts the routing database which provides routing
information in step 125 to the I-CSCF 612 identifying the
server 606". In step 126, the I-CSCF contacts the subscriber
independent server 606".
In the terminating case where the user is in a different
network to the subscriber independent server 606", the
following steps are carried out: the user 624 sends contacts
the P-CSCF 622 in step 131.. The P-CSCF 622 contacts the
associated S-CSCF 620 in step 132. These elements are outside
the network containing~the server 606".
46
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
In step 133, the I-CSCF 612, in the same network as the server
606" is contacted by the S-CSCF 620. Steps 124 to 126 are
then performed, as already described.
Where the user is in a circuit switched domain 654, the
circuit switched domain 654 contacts the MGCF 652 in step 141.
In step 142, the MGCF 652 contacts the I-CSCF 612. Steps 124
and 126 are then carried out as already described.
If a server handles all needed identities in its own database
or databases, it is not dependent on the HSS or any other
subscriber database. For this reason, it can be referred to
as a subscriber independent server. In preferred embodiments
of the present invention, the subscriber independent server
may not be an application server so an ISC interface may not
be used. It can be regarded as being similar to an entry
point to another network and can be looked on by the network
of which it is a part as if it were a I-CSCF of another
network. Subscriber independent server may be located
physically also outside the network. All the required data
concerning the subscribers is located in the server itself or
in its own database.
Embodiments of the present invention have the advantage that
routing becomes simple. No S-CSCF involvement is required in
the cases where the .group server is the originator of a
session or a transaction. No HSS involvement is necessary.
For example, the SLF can offer ari address to the server in the
terminating case. All data concerning the subscribers can be
in the database of the subscriber independent server or in a
database or databases connected to the server. The Sh i.e.
the interface between HSS and AS is not used. Sh interface may
be used to get the S-CSCF address from the HSS in the case
when an AS originates a session or transaction. If no S-CSCF
47
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
is involved in the routing there is no need to ask HSS the S-
CSCF address and thus no Sh interface is necessary.
Embodiments of the present invention have the advantage that
third party operators Can easily offer originating and
terminating services and there is no need to insert anything
into the HSS. The only Change which may be required is to
insert a domain name address pointing to the server in the
SLF. For example news.3-party-operators.operator.net may be
inserted to the SLF and connected to the routing to a
subscriber independent server,~located e.g. in the address
"news-host.newscompany.3-party-operators.operator.net" i.e. in
the subdomain 3-party-operators.operator.net. As well the
domain name may be completely different from the operator's
own domain name. For example the entry news.company.com may be
inserted to the SLF and associated to the routing to a
subscriber independent server e.g. a news server, of the
company.com. In this way a third party operator may be able
to offer services to IMS subscribers without having to have
its own IMS network. Thus embodiments of the invention, all
needed data relating to the subscribers may be located in the
server itself or in its own database or databases or in
databases operated by the same operator as the operator of the
server. This makes it possible for the third party to offer
services from its own server and to utilise the main (that is
a different) operator's IMS or similar network for routing.
Subscribers of the subscriber independent server may or may
not also be IMS subscribers. The third party operator is able
to run its server independently of the main operator.
In one modification to the embodiments described, the outbound
proxy is implemented by a S-CSCF so that the originating AS
sends a signal to the S-CSCF to act as an outbound proxy
instead of a S-CSCF. '
The signal sent by the AS is in the initial request
48
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
a) embedded in the address of S-CSCF e.g. it may be a
parameter, a port number, a Character or bit string in the
user part of the address and/or
b) as a separate signal from the address of S-CSCF e.g. in a
separate header or in the payload.
Because the outbound proxy is only a subset of functionalities
of S-CSCF it is simple to implement . With the same signalling
mechanism, outbound proxy can be implemented with I-CSCF too,
or with whichever CSCF.
The Release 6 version of the third generation partnership
standard 23.228, which is hereby incorporated by reference,
introduces the concept of a Public Service Identity (PSI). The
arrangement discussed below uses the S-CSCF arranged to
provide either a S-CSCF functionality or a outbound proxy
functionality.
With the introduction of standardized presence, messaging,
conferencing, and group service capabilities in IM ' CN
subsystem, there is a need for Public Service Identities
(PSIs). These identities are different from the Public User
Identities in the respect that they identify services, which
are hosted by application servers. In particular, Public
Service Identities are used to identify groups. For example a
chat-type service may use a Public Service Identity (e. g.
sip:chatlist X@example.com) to which the users establish a
session to be able to send and receive messages from other
session participants.
Public Service Identities take the form of a SIP URL as
defined in RFC 3261 [12] and RFC 2396 [13] or the "tel:"-URL
49
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
format as defined in RFC 2806 [15]. These standards are hereby
incorporated by reference and are IETF standards.
The IM CN subsystem provides the capability for users to
create, manage, and use Public Service Identities under
control of AS. It is possible to create statically and
dynamically a Public Service Identity. Each Public Service
Identity is hosted by. an application server, which executes
the service specific logic as identified by the Public Service
Identity. The IM CN Subsystem provides a capability of routing
IMS messages using Public Service Identity.
The routing of the AS originated sessions/transactions with
Public Service Identity is not clear in the current proposals
and the arrangements described below addresses this.
Up until now only the routing towards a PSI has been
described, i.e. requests that terminate at the AS that
provides the service. Embodiments of the present invention
discuss different possibilities for routing of requests that
originate from a PSI.
Requests originating from a PSI are required e.g. when a
Conference AS invites a user to a conference (dial-out). As
this example shows, the progress of the conferencing work in
CN1 is strongly related to the PSI routing procedures.
For routing of requests that originate from 'a PSI, the
following possible routing scenarios can be used:
a). Request always routed via a S-CSCF in the originating home
network
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
In this case the AS 700 always has to route via the S-CSCF 704
of its home network first.
This can be achieved by placing a so-called pre-loaded route
header into the request (standard SIP procedure). The' routing
is then from the S-CSCF 704 to the terminating I-CSCF.702,
b) Request always requires routing via any CSCF in the
originating network
Here the AS routes the initial request to either the I-CSCF
706 or the S-CSCF 704 of the home operator first. The
particular CSCF can be determined either dynamically (e. g.
over the Sh interface) or due to operators policy.
c) Request always routed directly to the destination network
The AS 700 in this scenario routes directly to the terminating
I-CSCF 702, without any involvement of a CSCF in the
originating network. This is also inline with the routing
procedures as described in SIP.
d) Request routed due to operator decision
Due to the possibility of having a pre-loaded route, it is not
required to standardize one of the above scenarios as the only
valid one for IMS - the routing behavior of the AS can be
determined by operator based on the policy of the home
network.
.Based on the provided service,, an AS may or may not support
specific routing functionalities. SIP provides the possibility
that an entity is only able to route to a dedicated next hop,
the so-called Outbound Proxy. If an AS is not able to e.g.
51
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
resolve the address of the terminating I-CSCF, it needs to
forward the request first to an entity that is capable of
routing the request towards the terminating network.
This especially might be the .case when the terminating party
is indicated by a tel URL. In order to resolve a tel URL the
AS could route the request first to the S-CSCF, which is able
to resolve tel URLs.
On the other hand it is very likely that many application
servers will be able to perform SIP routing procedures, DNS.
The functionality of the S-CSCF may need to be adjusted in
order to provide the necessary routing mechanisms for AS's;
the S-CSCF should perform only its routing capabilities (and
not e.g. the filtering capabilities), when it detects that an
incoming originating request indicates a PSI as the
originator.
Depending on the nature of services some charging support can
already be provided within the S-CSCF. Hov~iever, the charging
for specific services in IMS is not performed by the CSCFs, as
they are designed to be service agnostic. If the charging
support provided by the S-CSCF is not enough the AS can
provide more information for charging purposes.
Nevertheless, in the given example for a dial-out conference,
the invitation will also involve a media session between the
AS and the called user. In this case the generation of
charging information for the~session - based on the SDP in the
INVITE message - could be performed by the S-CSCF.
It has to be noted, that in this case, the S-CSCF would
52
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
a) need information about the user, to whom the PSI relates
to (e.g. conference creator) - the PSI itself does not
include any hint for the user who has to be charged;
b) not have any control over the media session itself (as
e.g. the P-CSCF/PCF has via the Go interface).
The operator might want to collect certain data from all calls
that traverse its network. Such functionality can be performed
by e.g. an I-CSCF, in order not to use too much of the
resources of the S-CSCFs.
As shown above, there might be cases where an operator wants
to route PSI originated calls to a CSCF in its own network
first, although SIP allows that the AS resolves the
terminating I-CSCF and routes to it directly.
It is also clear, that the routing behaviour may be different
for the individual cases which calls for a certain level of
flexibility in routing:
a) the operator might want to force all AS's to route PSI
originated calls over one or more specific entities in
its network'(strict policy);
b) the operator might want to force only certain AS's to
route PSI originated calls over one or more specific
entities in its network;
c) although the operator does not apply any routing policy,
the AS might not be able to perform SIP routing
procedures and therefore needs to contact the S-CSCF
first;
d) although the operator does not apply any routing policy,
the AS might need to contact the S-CSCF in certain cases,
e.g. when ENUM cannot be performed by the AS (case-by-
case routing);
53
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
Allowing such a flexible approach would on the other hand
would deviate from some principles within IM CN Subsystem as
it currently is, e.g.
a) if the operator applies a lose policy, the AS could route
directly to entities outside the home network, although
there is no interface defined for such purpose;
b) if the operator applies a lose policy, the AS could route
directly to the BGCF (e.g. when inviting another user to
a conference) ;
c) if the operator does not force the AS to route over the
S-CSCF, the S-CSCF might not get aware of the. media
streams that are originating from / terminating at the
home network;
d) the routing of calls originating from an AS / PSI would
not be strictly defined within the home network and based
on the individual case and operator policy, the routing
behavior will be different.
If the sessions/transactions are routed via the S-CSCF, the
first problem is what S-CSCF should be used and second how to
skip over the filter criteria handling. This is illustrated by
the arrangement shown in Figure 9.
The AS 800 fetches the S-CSCF address from configuration data.
The AS sends a first message to the identified S-CSCF 802:
INVITE X from Y ( Y is the PSI identity) Route:
psiscsf.home.net.
Because the Route contains the PSI indication, the S-CSCF 902
skips the evaluation/processing of the filter criteria. The S-
CSCF 802 then send INVITE X From: Y message to the I-CSCF 804.
Embodiments of the present invention have been described in
relation to application servers. However ~ it should be
54
CA 02516774 2005-08-19
WO 2004/075507 PCT/IB2004/000546
appreciated that embodiments of the invention can, also be used
with gateways or any other entity especially with an entity
having the same or similar relationship as the application
servers to other entities illustrated in the Figures and/or as
described.
It should be appreciated that a number of different features
have been described and that is possible that some embodiments
of the invention can combine different ones of these features.
It should be appreciated that in embodiments of the present
invention, IMS is access independent. This means that any
suitable access method such as WLAN (wireless local area
network) or the like can be used. IMS and PRES schemes offer a
way to specify services without specifying the protocol to be
used to get services. These protocol independent schemes
provide a way to identify services.
