Note: Descriptions are shown in the official language in which they were submitted.
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
METHOD AND APPARATUS FOR IMPROVED CALL SETUP IN A
MULTIMEDIA PACKET NETWORK
DESCRIPTION
Technical Field
This invention is related to multimedia packet networks. Specifically,
this invention relates to a mechanism to allow such a packet network to
more effectively carry telephony messages, and to more efficiently interface
with the public switched telephone network (PSTN).
Background Art
Evolution of the PSTN has accelerated in recent years; however,
most of the PSTN still operates on circuit switched, time division multiplexed
(TDM) connections. Integrated services digital network (ISDN) bearer
channels often provide transport. In parallel with the PSTN, a packet based
data network has evolved. This data network has largely been used for
Internet traffic and data networking. Although these networks have been
mostly separate until recently, the two networks are starting to merge. The
merger of these networks requires that voice traffic be carried over packet
networks, and further that such packet networks be able to seamlessly
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
integrate with traditional circuit switched networks, as the two types of
networks may carry different call legs of the same call.
FIG. 1 illustrates a typical TDM, PSTN call. Caller 101 places a call
to callee 105. The call goes through end office A, 102, over some type of
trunk bearer channel to toll office 103, then to end office B, 105, and
finally
to the callee. Such calls may pass through multiple toll offices, or may be
connected directly from one end office to another. In any case, a path of
circuits for the call is maintained throughout the call. Signaling between
offices is typically provided by an ISUP (ISDN user part) connection. ISUP
signaling is well understood and is standard in the telecommunications
industry. For more information on ISUP signaling, see the various Interna-
tional Telecommunications Union (ITU) Recommendations pertaining to
telephone signaling, including Q.761, Q.764 and Q.931, the most recent
versions of which at the time of filing this application are incorporated
herein
by reference.
FIG. 2 illustrates a call which is similar to the TDM call of FIG. 1;
however, in this case, the call is transported from one end office to another
(called switch offices, 202 and 204, in this case) via a packet switched
network 203. This fact is, in theory, transparent to caller 201 and callee
205.
ISUP+ or SIP+ provides signaling in this case. ISUP+ is essentially the
same as ISUP except that ISUP+ signals contain extra fields for packet or
cell based network information. An International Telecommunications Union
2
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
(ITU) recommendation was proposed for ISUP+ as of the filing date of this
application as ITU Q.BICC. SIP stands for "session initiation protocol" and is
a well-known standard. SIP and SIP+ are described in document RFC
2543, published by the Internet Engineering Task Force (IETF), March, 1999
which is incorporated herein by reference. SIP and SIP+ provide the same
type of signaling for control of calls, but are more oriented towards packet
based networks.
The network of FIG. 2 has been conceptualized for some time, and
standards groups and conference groups have written extensively about
how to make such a network work in everyday use. In order for the call leg
which is handled by the packet network to seamlessly connect with the call
legs handled by TDM switching offices, media provided by one type of
network must be converted into media provided by the other. This conver-
sion is referred to as circuit emulation services (CES) in an ATM network.
The device that provides this conversion is called a media gateway (MG). In
the network of FIG. 2, a media gateway handles each end of the bearer
connection through packet network 203. The media gateway terminates
bearer media streams from both the switched circuit TDM network, and the
packet network. The media gateway and the network it serves may be
capable of processing audio and video (hence the term "multimedia packet
network"). The media gateway is capable of full duplex media translations.
It may also provide other features such as conferencing.
3
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
Each media gateway is associated with a media gateway controller
(MGC). The media gateway is "dumb" in that it does not have call process-
ing capabilities. The call processing capabilities for the network reside in
the
MGC. An MGC provides the signaling for call control and controls the call
state of a media gateway. The protocol used by the MGC to control the MG
is called the media gateway control protocol (or the "Megaco" protocol).
Megaco is an application layer protocol which is also described in ITU
Recommendation H.248, which shares a common text with the IETF Internet
Draft "Megaco Protocol," and which is incorporated herein by reference.
The "Megaco Protocol" Internet Draft first became an IETF working docu-
ment in March, 1999. Within the Megaco protocol, session description
protocol (SDP) can be used to describe bearer channel terminations, which
are being controlled by the MGC's. SDP is described in document RFC
2327, published by the IETF, April, 1998, which is incorporated herein by
reference. Throughout the rest of this disclosure we will refer to Megaco as
"Megaco/H.248."
Despite the fact that the theoretical workings of a network like that
shown in FIG. 2 have been widely explored, such networks have seen
relatively little everyday use. The reason is that there are still problems to
be overcome before such networks exhibit the same very high quality of
service for voice traffic that users of the PSTN have come to expect. Once
such problem stems from the fact that there is no dedicated physical path for
4
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
a call through a packet network, and therefore a wait state must be provided
so that each node in the network waits for an entire bearer path to be estab-
lished before timing out.
A packet switched network, used for transport of audio and video
media streams, is typically based on asynchronous transfer mode (ATM)
frame relay (FR), and Internet protocol (1P) technologies. Public ATM
networks generally operate according to the user network interface (UNI).
The UNI is described in the book, "ATM User Network Interface (UNI)
Specification Version 3.1" by the ATM Forum, published by Prentice Hall
PTR, June, 1995, which is incorporated herein by reference. An update to
the UNI version 3.1, "ATM User-Network Interface (UNI) Signaling Specifi-
cation 4.0" was published by the ATM Forum in July, 1996, and is incorpo-
rated herein by reference. For private ATM networks, the private network to
network interface (PNNI) describes the ATM interface. PNNI is covered in
the ATM forum document "PNNI addendum for the network call correlation
identifier" published by the ATM forum in July 1999, which is incorporated
herein by reference.
In ATM, fixed-length cell carry packetized data. Each cell that is sent
through the network has a virtual channel identifier, and other addressing
information; however, each node in the network handles many cells that are
associated with different media streams. Therefore, there is no provision for
a node to control cells for a call throughout a call path. Media gateways and
5
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
other nodes in the path must actively wait for call setup throughout the
network, consuming processing power and making it difficult to maintain an
appropriate level of quality of service. What is needed is a way to more
directly monitor the status of a connection, so that valuable processing
power is not used to process wait states.
Disclosure of Invention
The present invention solves the above described problem by pro-
viding new signals and events to be used in call setup of a connection
oriented bearer path so that no wait state is required during the path setup.
These signals and events allow processors in the network to switch to other
tasks instead of actively waiting for connections to be completed. The
signals and events also provide for simpler and more efficient network
implementation. The invention can be used exclusively to control call setup.
Or it can be used in combination with transaction pending messages.
A media gateway and its MGC at either or both ends of a bearer path
can make use of the invention. In describing the invention, we use "originat-
ing" and "terminating" to refer to the calling and called ends of the call
path,
respectively. We use the terms "near-end" and "far-end" to refer to the end
of the path relative to where the particular process being discussed is taking
place, usually relative to where key messages are being exchanged by a
media gateway and an MGC. The terms "near-end" and "far-end" are used
6
CA 02383345 2002-02-21
WO 01/15486 PCTNS00/22853
independently of the terms "originating" and "terminating." In connection
with call setup, the terms "forward" and "backward" refer to which end
initiates the bearer connection through the packet network. "Forward" refers
to a process where the originating end sets up the connection, and "back-
ward" refers to a process where the terminating end sets up the connection.
In the Megaco/H.248 protocol, a signal is a request from the MGC to
the media gateway to apply some process to network terminations. In the
case of TDM terminations, the signals can be in-band signals, such as
tones, or out-of-band signals such as those specified by ISUP. In any case,
these signals are relayed from the media gateway to the MGC where they
are processed. In the case of ATM terminations, most signals are UNI
signals or messages. An event is a notification sent from the media gate-
way to the MGC when a requested state is reached. In the preferred em-
bodiment, the invention is implemented with new events and signals to be
used in the Megaco/H.248 protocol. However, the fundamental operation of
these events and signals is independent of the underlying bearer network.
The signals and events can be mapped to any underlying network such as
ATM or frame relay.
According to the invention, a media gateway receives a command
from its MGC to add a connection for the bearer path and the command
specifies a connection available signal for notifying the MGC if and when the
connection is available. We abbreviate "connection available" as "coav."
7
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
Coav is both a signal and an event. When an ATM network is used for
transport, the command from the MGC triggers the media gateway to send
an ATM UNI setup message to the far-end media gateway. The media
gateway then initializes the connection with the far-end media gateway. If
there is no connection at all for the call, this initialization step involves
setup
and connect messages being exchanged. In most cases an "add" command
is used to direct a media gateway to establish a bearer path, however, some
other command, for example, a "move" command can be used. Also, as will
be explained in more detail later, there could be a connection which has
already been established by the far-end media gateway, in which case, the
initialization step simply involves identifying the connection and correlating
it
with the current call. Once the bearer path has been established, the coav
event is sent to the MGC.
If the connection cannot be established, a "connection not available"
or "cont" signal is specified. Like coav, cont is both a signal and an event.
The MGC can also send cont to the media gateway to disconnect or release
the path. The media gateway in this case can respond with the connection
not available event. The media gateway will notify the MGC when the
connection is disconnected at any time during the call, even if the discon-
nection occurs because of a network failure. In this case, a notice of failure
event may also be sent from the media gateway to the MGC.
8
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
As part of the process described above, an optional continuity check
can be performed across the bearer path. This check involves one of the
media gateways sending a continuity check message to the other media
gateway. The media gateway receiving the continuity check then responds
with a continuity response. In this way, a bearer path can be confirmed at
the end of the setup process to ensure the reliability of the call being han-
died.
The invention is implemented by software in combination with the
hardware of the media gateway and media gateway controller. The software
which implements many aspects of the present invention can be stored on a
media. The media can be magnetic such as diskette, tape or fixed disk, or
optical such as a CD-ROM. Additionally, the software can be supplied via a
network. A media gateway is essentially a switching system containing
switching fabrics, a computing module, network interfaces, and other re-
sources. The network interfaces are implemented by adapters which are
connected to switching fabrics to allow access to the system from the net-
works. Input/output modules or adapters allow software to be loaded and
various maintenance functions to be performed. A computing module
contains a processor and memory that execute the software and provide the
means to control the operation of the media gateway to implement the
invention.
9
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
The media gateway controller can also be a switching system, but
would more typically be a type of workstation containing a bus such a
personal computer interconnect (PCI) bus. A workstation that typically
implements the invention includes a plurality of input/output devices and
adapters for connection to the necessary networks. A system unit includes
both hardware (a central processing unit and memory) and software which
together provide the means to implement the media gateway controller.
The invention operates in a network in which media gateways act as
endpoints to a call leg being carried on a bearer channel through the net-
work. Each media gateway is controlled by and connected to a media
gateway controller. An MGC uses the previously mentioned Megaco/H.248
protocol to control its media gateway, and the invention provides an exten-
sion to the Megaco/H.248 protocol.
Brief Description of Drawings
FIG. 1 conceptually illustrates a prior-art telephone connection
through the public switched telephone network.
FIG. 2 conceptually illustrates a telephone connection similar to that
of FIG. 1, except that one call leg goes through a packet switched network.
FIG. 3 is a block diagram of one network in which the present inven-
tion is used.
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
FIG. 4 is a block diagram of a different network in which the present
invention is used.
FIG. 5 is a flowchart illustrating the method of the present invention.
FIG. 6 is a flowchart illustrating the method of the present invention in
which a transaction pending message is used.
FIG. 7 is an example message flow diagram that illustrates an exam-
ple of how the present invention is used.
FIG. 8 is another example message flow diagram that illustrates an
example of how the present invention is used..
FIG. 9 is another example message flow diagram that illustrates how
the present invention is used.
FIG. 10 is another example message flow diagram that illustrates the
use of the present invention.
FIG. 11 is another example message flow diagram that illustrates the
use of the present invention.
FIG. 12 is another example message flow diagram that illustrates the
use of the present invention.
FIG. 13 is a block diagram of a media gateway that implements the
present invention.
FIG. 14 is drawing of one implementation of a media gateway con-
troller that is used with the present invention.
11
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
FIG. 15 shows an example of a media which stores software that
implements the present invention.
Mode ~ for Carrying Out the Invention
FIG. 3 illustrates one architecture in which the present invention can
be used. According to FIG. 3, telephone 304 is where a call originates.
Telephone 303 is where a call terminates. Telephones 303 and 304 are
shown as illustrations only. In reality, they can be directly connected to the
media gateways or can be connected through extensive TDM networks. In
the latter case, lines going into the media gateways would actually be TDM
trunks. Media gateway A, 310, is the originating media gateway and media
gateway B, 309 is the terminating media gateway. The media gateways of
FIG. 3 convert voice to ATM. We therefore refer to this network architecture
as voice and telephony over ATM, or "VTOA" architecture, however, the
same architecture can be used for Internet protocol (1P) telephony. Media
gateway controller A, 301, controls media gateway A. Media gateway
controller B, 302, controls media gateway B. Alternatively both media
gateways can be controlled by a single MGC. Media gateway A includes
TDM endpoint 305 and packet endpoint 306. Media gateway B includes
TDM endpoint 308 and packet endpoint 307. Packet network 311 serves as
the transport network through which bearer channels are established to
interconnect calls between the two media gateways. This network and the
12
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
endpoints to which it is connected can be frame relay, ATM, IP, or some
other type of packet network. For illustrative purposes, most of the discus-
sion refers to an ATM network. The media gateway controllers communi-
cate with each other via ISUP+, SIP, or SIP+. It is also possible to use a
nonstandard protocol, specific to the manufacturer of the media gateway
controllers and media gateways.
Either a media gateway or a media gateway controller can generate
the end-to-end call identifier (EECID), as determined by the network de-
signer. The EECID is used to identify a call leg uniquely across the ATM
network, regardless of the number of nodes used in completing the network
path. The EECID allows the MGC's, the media gateways, and any nodes in
the bearer path to identify the call uniquely. Note that media gateway
controller A, 301, controls media gateway A, 310, using the Megaco/H.248
protocol, an application layer protocol for media gateway control. Likewise,
media gateway controller B, 302, controls media gateway B, 309, using the
same Megaco/H.248 protocol. The media gateway or media gateway
controller at either end can generate the EECID, regardless of which end is
the originating end for the call and which end is the terminating end for the
call.
FIG. 4 illustrates a slightly different architecture in which the invention
is used. According to FIG. 4, media gateway controller A, 401, controls
media gateway A, 415, using the Megaco/H.248 protocol and media gate-
13
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
way controller B, 402, controls media gateway B, 414, using the
Megaco/H.248 protocol. In FIG. 4, 403 is the originating telephone and 404
is the terminating telephone. ATM network 411 serves as the transport
network. Again, any of the media gateway controllers or media gateways
can generate an EECID to identify calls being handled by the network. The
main difference between the network of FIG. 4 and the network of FIG. 3 is
that the network of FIG. 4 supports digital subscriber loop, or DSL. DSL
comes in various types such as aDSL, sDSL and hDSL, and so "xDSL" is
used to designate DSL in FIG. 4. In this case each media gateway includes
a splitter; 405 in the case of media gateway 415 and 410 in the case of
media gateway 414. TDM terminations 406 and 408 and ATM endpoints
407 and 409 each reside in their respective media gateways and allow both
data and TDM voice to be transported across the ATM network 411. The
splitters 405 and 410 split the voice from the data. The data connection
from user terminal 412 is completed through splitter 405 to ATM termination
407 in the case of media gateway A. The data connection from user termi-
nal 413 is completed through splitter 410 to ATM termination 409 in the case
of media gateway B. Otherwise, the operation of the network in FIG. 4 is
essentially the same as the operation of the network of FIG. 3.
Many aspects of the invention are implemented through enhance-
ments to the previously mentioned Megaco/H.248 protocol. The connection
model for the protocol describes logical entities, or objects, within the
media
14
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
gateway that can be controlled by the media gateway controller. The model
relies on extractions, primarily terminations and contexts. A termination
sources and/or sinks one or more media streams. A context is an associa-
tion between a collection of terminations.
In general, an "add" command is used to add terminations to con-
texts. A termination may be moved from one context to another with a
"move" command. A termination exists in, at most, one context at a time. A
non-packet termination can exist outside of a context. Property values can
be set for terminations by including appropriate descriptors as parameters to
the various commands in the Megaco/H.248 protocol. A termination in a
context may have its value changed by the "modify" command. Other
commands that are important to the implementation of the invention will be
discussed later.
As previously mentioned, according to one aspect of the invention an
end-to-end call identifier (EECID) is associated with a call, and with a
bearer
path through the packet network, which completes a call leg. When we say
the EECID is associated with a call or a path, we mean that all of the nodes
and devices involved in maintaining a call leg are aware of which call to
which specific Directory Numbers (DN's), or other user address are associ-
ated with each packet of information which flows through the relevant part of
the network. Depending on the type of underlying networks and/or protocols
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
the EECID can be carried across the network in various ways. Details of
some possible signaling will be discussed later.
It is preferable to include the EECID in the Megaco/H.248 protocol as
part of the stream descriptor in addition to the local control descriptor,
local
descriptor, and remote descriptor. These descriptors are all part of the
stream parameter, a known part of the Megaco/H.248 protocol. It is also
possible to include the EECID in the Megaco/H.248 protocol as part of a
session descriptor protocol (SDP) term. SDP is a well-known protocol,
described in the previously cited IETF RFC 2327, which is used to describe
packet terminations, such as IP and ATM terminations within the
Megaco/H.248 protocol.
In addition to including the EECID in the Megaco/H.248 protocol, it
must be included in other protocols and/or data streams that allow the
network to communicate. It is especially important to include the EECID in
the ATM cell structure used in the ATM transport network, since the media
gateways on the ends of the ATM networks form the ends of the bearer
channel carrying the part of the call leg which passes through the packet
network. Assuming the packet network shown in FIG. 3 and FIG. 4 is an
ATM network implemented according to the UNI standard promulgated by
the ATM forum, there is a choice of possible places in an ATM cell where
the EECID can be placed. The network prefix is a fixed, required part of the
cell, used for routing. The EECID could be placed in the ATM user part.
16
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
ATM network routing only uses the first thirteen-byte network prefix of the
ATM address. The following 7 bytes of the user part can be used to trans-
port the EECID. Another possible place for the EECID is the ATM subad-
dressing. The subaddressing field usually only has local significance and
can be dropped if it is unused. It can be adapted to implement the EECID of
the present invention. Most non-UNI 4.0 compliant ATM networks are
currently implemented without using a generic information trans-
port/information element (GIT IE) field; however, the GIT IE field will proba-
bly be the best place for the EECID as that field becomes more widely used.
The EECID must also be included in ISUP+ messages, if ISUP+ is
used between the two gateway media controllers. FIG. 8 shows an ISUP+
message. Field 801 contains the ISUP information and 802 contains the
ISUP+ information, which is essentially directed towards packet based
networks, and includes an application transport mechanism field. The
EECID is preferably included in the transport mechanism field.
If SIP+ is used between the two media gateway controllers, the
EECID is carried as a term in the session description protocol (SDP)
Syntax for the session description protocol, which includes the EECID, is
"c=eecid: (eecid value)" or "a=eecid: (eecid value)." Note that the terms
"c=eecid: (eecid value)" or "a=eecid: (eecid value)" need not be used if the
stream descriptor is used to specify the EECID in the Megaco/H.248 proto-
col.
17
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
As previously discussed, either an MGC or a media gateway some-
where in the network selects the EECID, that is, determines a value for the
EECID, before it can be associated with a call leg. The EECID can be any
arbitrary number that is unique so as to allow correlation of the end-to-end
network path between the two media gateways. The choice of the value for
the EECID has implications for the call flow. In some cases, the value can
only be derived by the network, as with the NCCI as discussed above.
Preferably, the value of the EECID is not dependent on the underlying
network architecture. A simple way to create an EECID is to simply have
the device that is determining the EECID, generate a random number. It is
also possible to use a number that is already associated with some part of
the network.
A possible value to use for the EECID is a session-ID or call-ID. The
session-ID is a random number passed from the MGC to the media gate-
way. The media gateway can then pass the session-ID to the far end media
gateway as an EECID with its initial setup message. The session-ID can
also be passed through ISUP+ messages. The session-ID would not be
able to be used if the media gateway is to generate the EECID. The call-ID
is similar to the session-ID. Both are specified to identify a call solely
within
an MGC or media gateway.
Finally, the most preferable value for the EECID, assuming a numeri-
cal value, which is associated with the network, is used, is the ATM sup-
18
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
ported backward network connection identifier (BNC-ID). The BNC-ID is
four bytes long and is generated by the media gateway. The media gateway
sends the BNC-ID to its media gateway controller for forwarding to the far-
end. The BNC-ID is included in the setup command between media gate-
ways to correlate the call.
According to our invention, a package with a signal, called a "connec-
tion available" (coav) signal, that explicitly requests the establishment of
the
packet network path, an event, called a "connection available" (coav) event,
that explicitly reports the successful completion of the path, and an event,
called "connection not available" (cont), that explicitly reports the failure
to
establish the requested path can be added to any protocol that is used for
call control in any packet-based network. Similarly, the package would
include a signal, called a "connection not available" (cont) signal, that
explic-
itly requests the release of the packet network path, and an event, called a
"connection not available" (cont) event, that explicitly reports the
successful
release of the path. If the invention is used with the Megaco/H.248 protocol,
these signals and events can be used alone, or with an existing provisional
response mechanism implemented through the "transaction pending" com-
mand. The events and signals that are used with Megaco/H.248 in both of
these alternatives are shown in the following table. The signals and events
symbolized by the "coav" and the "cont" symbols are new according to the
invention. The EECID, previously discussed, is an optional parameter for
19
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
the coav and cont events, hence it is denoted with the letter O. If a media
gateway fails to release a packet network path, the media gateway sends a
"report failure" (of) event to the media gateway controller. The continuity
check, continuity response, and report failure are also part of this package,
which we call the "packet pipe" event package:
Symbol Definition R S Parameters)
coav bearer connection availableX BR EECID (O)
cont bearer connection not X BR EECID (O)
available
cot continuity check X TO
cot continuity response X TO Duration
of report failure X Duration
R specifies that each symbol is part of an event report. BR indicates a brief
tone. TO indicates a timeout tone that stops after the amount of time speci-
fled by the duration parameter has passed. Note that the cot and cot
signal/events are shown for illustrative purposes only to demonstrate the
optional use of continuity testing in conjunction with the process of setting
up
a bearer path. These signals are not required to implement the present
invention in Megaco/H.248.
The explicit request alternative has desirable characteristics. In par-
ticular, the use of explicit signals and events eliminates the need for the
media gateway to maintain the state of an add transaction request. The
explicit embodiment also reduces the transaction request state monitoring in
the MGC, and eliminates the need for the media gateway to potentially send
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
multiple transaction pending replies. The explicit signals and events also
reduce complexity when multiple add commands are used in a single trans-
action.
When requesting establishment of a network path, an add command
is sent to the media gateway, which then explicitly specifies the "connection
available" (coav) signal and event. When requesting release of a network
path, a subtract command is sent to the media gateway, which then explic-
itly specifies the "connection not available" (cont) signal and event. The
coav signal is sent only to the ATM termination, which is responsible for the
origination of the setup sequence of the network path. The transaction,
which specifies these requests, is acknowledged on receipt. The media
gateway manifests the coav signal for an ATM network. A notify message of
the coav or the "of' event is sent from the media gateway to the MGC upon
a connection becoming available or upon failure to establish the connection,
respectively. Note that the use of the coav and cont signals are optional. If
they are not used, the initiation of the establishment or release of the path
is
implied by the add or subtract command, respectively.
The continuity check and response can be used automatically by the
media gateway without an instruction from the MGC. However, these two
events/signals can also be requested by the MGC during call processing.
An additional characteristic of this approach is that embedded signals and
21
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
events can be used to allow for additional processing to be invoked auto-
matically for such things as continuity checking of the network path.
An alternative way of making a connection request is based on the
Megaco/H.248 provisional response "transaction pending" reply wait state.
This command is used when a command is received but pending for com-
pletion of processing. The media gateway can respond to the MGC with a
command "transaction pending" response, so that the MGC won't be
blocked for the completion response. When the media gateway finishes
executing the command, it can then send the "transaction reply" message to
acknowledge that the original command has been successfully completed or
has failed. In addition to using the coav signal explicitly, the connection
request is expressed implicitly by the add command. The rationale behind
this approach is that the packet connection does not exist until it is added
to
the context. Therefore, the add command implies setting up the bearer
connection.
The EECID is present in either the stream descriptor or the termina-
tion descriptor for the network path. The media gateway must use the
EECID to determine if it needs to initiate the network path setup. The media
gateway will keep a record of all requests received from other media gate-
ways for setup of a network path. When the add command is received, the
media gateway will determine if a bearer path setup request with the speci-
fled EECID has been received. If a network path associated with the EECID
22
CA 02383345 2002-02-21
WO 01/15486 PCTlUS00/22853
exists, then the network path bearer connection already exists and the
correlation is reported back to the MGC via a transaction reply. If no pend-
ing network path is found with the same EECID, then the path initialization is
invoked. In this approach, there won't be a coav signal and bearer connec-
tion available event notification. If the media gateway determines that there
will be sufficient delay setting up the bearer connection to cause the trans-
action request to time out, the media gateway will respond to the media
gateway controller with a transaction pending response. Upon completing
the bearer connection, the media gateway will respond to the MGC with a
transaction reply indicating success or failure of the attempt.
The method described immediately above can incur processing over-
head in determining whether or not network path setup is required. Another
negative consideration is that there is no mechanism to use embedded
signals and events to allow for automatic processing of subsequent actions
such as continuity checking of the network bearer path. The MGC has to
issue a separate message for continuity checking and response. This
combination at least eliminates the need for the media gateway to search
through pending network path requests to determine if a network path setup
is required. If the coav signal is present, the setup will begin immediately.
If
the two embodiments are used simultaneously, accommodations have to be
made to eliminate redundant messaging to report the completion of the add
command and the coav or cont event.
23
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
Figures 5 and 6 illustrate the overall method of the invention. FIG. 5
is a flowchart that shows the explicit call setup method. At 501 the add
command is sent from the MGC to the media gateway. The add command
can specify any signal and response that the media gateway will under-
stand. In this case, the signal and response is coav for connection avail-
able, and/or cont for connection not available. At 502 the add command is
confirmed. In this embodiment, the confirmation takes the form of an imme-
diate transaction reply message to the MGC. At 503 the bearer connection
is initialized.
The bearer connection can be initialized either by the media gateway
setting up the bearer connection, called a forward setup, or the media
gateway simply identifying an existing bearer connection, which has been
set up by a far-end media gateway, called a backward setup. In either case,
the media gateway waits until the bearer path is completely established at
504, and responds with a coav event at 505. The MGC will typically notify
the far-end MGC of this occurrence at 506. An optional continuity check is
shown as being performed at 507. This continuity check consists of a
continuity check message being sent to the far-end media gateway, and a
response being received. If the connection is not available in FIG. 5, a cont
event is sent to the MGC at 508, and the far end is notified at 509. The
process then ends at 510.
24
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
The flowchart of FIG. 6 illustrates the implicit call setup method used
with the new signals and events of the invention. At 601 an add command
is given by the MGC to the media gateway as before. This time, however,
the media gateway responds with a transaction pending message at 602,
which forces the MGC to wait until it receives further instructions from the
media gateway. At 603, the media gateway again initializes the bearer path.
Only after the bearer path is fully established is the transaction reply sent
to
the MGC at 608. A coav event is returned to the MGC at 609. The far-end
is notified at 610. Again, a continuity check is optionally performed at 611.
In FIG. 6, if the bearer path has not been established at 604 after a
time-out at 605, the cont event is returned to the MGC at 606. The far end
is notified at 607. In this case, a continuity check can still be performed at
611 to further check the status of the network.
To illustrate the detail of the invention, Figures 7-12 present detailed
signal flows showing the setup of bearer path connections in a multimedia
packet network. There are literally dozens of possible signal flows, which
could be implemented to make use of the invention. The signal flows pre-
sented here should be considered as examples only. When we refer to
implicit versus explicit setup, we are using the terminology discussed above
for explicit versus implicit signaling and events. When we refer to forward
setup versus backward setup we are referring to which end of the network is
performing the bearer path setup relative to the originating end of the net-
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
work. If the originating end of the network is also setting up the bearer
connection we have a forward setup. If the originating end of the network is
passing information to the terminating end and the terminating end is setting
up the bearer connection, we have a backward setup. In reference to FIG.
9, all messages are discussed. For the other message flow diagrams, only
new messages, which are important to illustrating the differences between
those examples and previous examples, are discussed. The letters A and B
correspond to the ends of the network path as shown in the network dia-
grams of Figures 3 and 4.
In Figures 7-12, the TDM termination is a logical representation of a
TDM line and an ATM termination is a logical representation of an ATM
network connection. Although an ATM termination is illustrated in all cases,
the invention is not limited to use of an ATM network for the bearer connec-
tion. The invention is also applicable with other connection-oriented net-
works such as frame relay networks.
Turning to FIG. 7, an explicit forward setup is illustrated. At 710, a
notify message indicating an offhook condition is sent from media gateway A
to MGC A. At 711, MGC A responds with an add command. At 716, media
gateway A replies with a transaction reply. At 713 and 701, the two media
gateway controllers negotiate connection parameters. At 714, MGC A
sends a pipe connect request to MGC B. In this case, at 702, MGC B sends
an add command to media gateway B with explicit instructions for setting up
26
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
the bearer path with the connection signal coav and the event coav when
the connection is available. At 703, media gateway B immediately responds
to MGC B with a transaction reply signal; the transaction reply signal is in
response to an add command. This transaction reply does not mean that
the add command is completed. Rather, the transaction reply simply means
that media gateway B is working on adding the ATM termination. Media
gateway B chooses an EECID at 702 and sends the EECID back to MGC B
at 703. MGC B passes the EECID to MGC A at 707. At 715, MGC A sends
the add command with the EECID and explicit event and signal coav. Media
gateway A immediately replies to MGC A with a transaction reply at 717. At
704, the UNI setup message is sent from media gateway A through the ATM
network to media gateway B. A connect message is sent from media gate-
way B to media gateway A to indicate the bearer path is accepted at 705.
Media gateway B uses the EECID to associate the call and the bearer path.
This prevents an unauthorized bearer connection from being set up. Then
MGC B notifies media gateway B at 706.
After receiving a UNI connect message from media gateway B, media
gateway A notifies MGC A that the coav event has occurred at 708. In the
above example, the MGC B cannot add the ATM termination until the UNI
service has been set up. This limitation comes about because the EECID is
needed to create the ATM termination. At 718, MGC B is notified by MGC A
through ISUP+ or other means that the bearer path (packet pipe) has been
27
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
established. The process is completed with the pipe connect complete ack
message at 712.
FIG. 8 illustrates the signal flows for explicit backward setup. At 801,
MGC A chooses the EECID and passes it to media gateway A. There is an
embedded continuity check applied after the coav event occurs and there is
a return continuity check associated with the continuity check event. At 812,
MGC A passes the EECID to MGC B. At 802, media gateway A sends a
transaction reply to the add command to acknowledge that the transaction is
accepted. Once again, the add command has not been fully executed. At
811, MGC B sends an add command to MG B. This command asks for a
bearer path to be set up using signal=coav. At 803, media gateway B sends
the ATM UNI setup message with an EECID to media gateway A. An event
notification on coav with embedded event cot and signal cot is explicitly
requested. Upon receipt of continuity check cot response cot will be given.
This command also asks for a bearer path to be set up using signal=coav.
After the connection is set up, media gateway B responds at 805 with a coav
event. At 809 MG A also notifies MGC A that event=coav has occurred. At
806, the continuity check signal is applied by media gateway A since the
coav has occurred. At 807, media gateway B applies the continuity check
response signal since it receives the continuity check event. At 808, media
gateway B notifies its media gateway controller that the continuity check
event has occurred. Similarly, media gateway A notifies its media gateway
28
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
controller that the coav and the continuity check return events have both
occurred at 810.
FIG. 9 illustrates the message flows for implicit backward setup. In
this case, the EECID is assigned by the media gateway. At 901, media
gateway A chooses the EECID and sends a transaction pending response
with the EECID to its media gateway controller so that the media gateway
controller waits for the setup. At 902, media gateway B passes the EECID
in the UNI setup message to media gateway A. At 903, when the connec-
tion setup is complete, media gateway A sends a transaction reply to the
add command and the EECID to media gateway controller A. Media gate-
way B immediately sends a transaction reply to the add command to MGC
B.
FIG. 10 also illustrates an implicit forward setup. In this case a media
gateway controller creates the EECID. Also, the cot and cot continuity
checking messages are used. At 1001, an add command with the EECID is
sent from media gateway controller A to media gateway A. At 1002, media
gateway A responds to its MGC with a transaction pending command. Note
that media gateway A still uses the EECID for establishing the connection
with media gateway B. 1003 and 1004 illustrate the continuity check and
continuity check response, respectively.
Figures 11 and 12 illustrate what happens when a failure occurs. In
FIG. 11, at 1101, MG B can't accept the UNI setup due to an error. Any
29
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
number of things could cause the error. One possibility is that there is no
EECID known at MG B that matches the one in the UNI setup message. At
1104, MG B reports failure (of) to MGC B. At 1102 MG A times out or
receives a reject message from MG B. MG A reports failure (of) to MGC A
at 1103. At 1105, and 1106, MGC a and MGC B exchange messages to
disconnect the pipe connection. Messages like that shown at 1105 and
1106 can come from either MGC A or MGC B.
In FIG. 12, the EECID is created and sent from MG A to MGC A at
1201. MG A sends the UNI setup message at 1202, but it is rejected at
1206. After the add command is confirmed at 1205, a report of failure (of) is
sent from MG A to MGC A at 1208. A report of failure is also sent from MG
B to MGC B at 1207.
FIG. 13 illustrates a conceptual, functional block diagram of a switch-
ing system, which can be used to implement a media gateway, which in turn
implements the invention. Computing module 1301 includes a central
processing unit, memory, and supporting circuitry. This computing module,
together with any computer program code stored in the memory, is the
means for controlling the overall operation of the switching system to per-
form the method of the invention. TDM switching fabric 1302 is for switching
TDM channels and is controlled by the computing module. Input/output (I/O)
module 1304 is also connected to the processor of computing module 1301
and includes media devices to load computer program code as well as
CA 02383345 2002-02-21
WO 01/15486 PCTNS00/22853
connections for workstations or other equipment for control and mainte-
nance of the switching system. TDM network access module 1303 serves
as a TDM network interface and is connected to TDM switching fabric 1302,
both of which are managed by the computing module 1301. Circuit emula-
tion system 1305 provides circuit emulation services, converting TDM voice
to packets such as ATM cells. Packet switching fabric 1306 sends and
receives packets on the packet network through packet network interface
1307.
FIG. 14 illustrates a workstation, which can be used to implement a
media gateway controller according to the present invention. I/O devices
such as keyboard 1402, mouse 1403 and display 1404 are used to control
the system. One or more of these devices may not be present in normal
operation. System unit 1401 is connected to all devices and contains
memory, media devices, and a central processing unit (CPU) all of which
together form the means to implement the present invention. Network
interfaces are normally implemented via adapter cards plugged into a bus,
however, for the sake of simplicity they are shown graphically as interface
1405.
As previously mentioned, appropriate computer program code in
combination with appropriate hardware implements most of the elements of
the present invention. This computer program code is often stored on
storage media. This media can be a diskette, hard disk, CD-ROM, or tape.
31
CA 02383345 2002-02-21
WO 01/15486 PCT/US00/22853
The media can also be a memory storage device or collection of memory
storage devices such as read-only memory (ROM) or random access mem-
ory (RAM). Additionally, the computer code can be transferred to the work-
station over the Internet or some other type of network. FIG. 15 illustrates
one example of a media. FIG. 15 shows a diskette of the type where
magnetic media 1502 is enclosed in a protective jacket 1501. Magnetic field
changes over the surface of the magnetic media 1502 are used to encode
the computer program code. In this way the computer program code is
stored for later retrieval.
We have described specific embodiments of our invention, which pro-
vides an end-to-end call identifier (EECID) to uniquely identify a call leg
across a packet network, regardless of the number of nodes used in com-
pleting the network path. One of ordinary skill in the networking and com-
puting arts will quickly recognize that the invention has other applications
in
other environments. In fact, many embodiments and implementations are
possible. The following claims are in no way intended to limit the scope of
the invention to the specific embodiments described.
We claim:
32
