Note: Descriptions are shown in the official language in which they were submitted.
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
CIRCUIT INTEGRITY IN A PACKET-SWITCHED NETWORK
BACKGROUND.
The invention relates to circuit integrity in a packet-switched network.
System Signal 7 (SS7) messages are often used to provide control signals in
various telecommunications systems, such as telephone systems, and provide a
mechanism, known as continuity check, for checking the integrity of a circuit
between
two switching network endpoints during call setup. Continuity checks
originally were
developed for analog facilities and consist, for example, of a frequency tone
transmitted by the originating exchange and looped back by the receiving
exchange.
Reception of the returned tone by the originating exchange indicates that the
channel
is available. In digital environments, use of continuity check operations has
been
similar.
Recently, packet-domain network architectures, such as asynchronous transfer
mode (ATM) networks, have been considered for transporting voice and other
narrowband traffic. Packet networks allow connections to be made between
endpoints without dedicated inter-switch connections. Fixed-size packets of
data,
known as cells, are transferred between the ATM switches, which are packet
switches
that provide virtual circuits between the end points of a network.
With the advent of packet voice networks and the introduction of adaptation
between circuit-switched and packet-switched bearers, high-quality integrity
checks
should be capable of detecting and isolating various types of failures.
Traditional
testing by continuity check tones, however, is incapable of detecting certain
failures
such as, for example, impaired bits having low significance with respect to a
time-
domain multiplexed (TDM) sample value. Therefore, better-quality continuity
checking suitable to packet networks is desirable.
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
SUMMARY
In general, according to one aspect, techniques for performing a continuity
check operation include sending a pattern of bits over a packet network
connection
through a first interface on a packet-switched network to a second interface
on the
packet-switched network. The first interface is monitored for return of the
pattern of
bits over the packet network connection. A decision whether the continuity
check is
successful is based on whether the pattern of bits is detected at the first
interface
during the monitoring.
In various implementations, one or more of the following features may be
present. The technique can be used in connection with both narrowband and
broadband communications. The continuity check can be performed, for example,
during a set-up process for a narrowband call over a packet network. The call
set-up
process can include sending Signaling System 7 (SS7) messages. The techniques
can
include providing a loop that connects incoming and outgoing packet streams
associated with the packet network connection. The particular pattern of bits
may
vary depending on various factors including the type of failures to be
detected. The
pattern of bits can be sent repeatedly over the packet network connection
during the
monitoring.
A system for performing continuity check operations for traffic that is to be
transported over a packet network also is disclosed. The system can include a
first
gateway that is coupled to a first interface on the packet network and that is
configured to execute continuity check operations. The gateway includes a bit
pattern generator arranged to generate a pattern of bits to be sent over a
packet
network connection, and a bit pattern detector arranged to monitor return of
the
pattern of bits over the packet network connection. The gateway is configured
to
decide whether a continuity check is successful based on whether the generated
pattern of bits is detected by the bit pattern detector.
The system also can include a second gateway coupled to a second interface
on the packet network and configured to provide a loop between incoming and
outgoing packet streams associated with the packet network connection. In some
2
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
implementations, the gateways may be configured to adapt circuit-switched and
packet-switched bearers.
Various implementations may include one or more of the following
advantages. The techniques described here can help ensure the non-disruptable
transfer of data through a packet-switched network. In particular, continuity
check
operations employing a pattern of bits can be used to test for and isolate a
wide range
of potential failures that may occur in a packet-domain connection or the
adapters
terminating the path. Therefore, the continuity checks can help ensure that
the circuit
will faithfully reproduce the signals traversing it. In some cases, a pattern-
based
continuity check operation can reduce the number of resources required to
execute the
test compared to frequency or tone-based continuity check operations.
Other features and advantages will be readily apparent from the following
detailed description, the accompanying drawings and the claims.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of a telephone connection through a hybrid ATM
network and an associated signaling network.
FIG. 2 is a simplified block diagram of an exemplary media gateway.
FIGS. 3A and 3B are a flow chart of a method for employing continuity
checks in a voice call set up process over an ATM network.
FIG. 4 is a signal flow diagram corresponding to FIG. 3.
FIGS illustrates further details for performing a continuity check operation.
FIG. 6 illustrates an exemplary bit pattern.
DETAILED DESCRIPTION
As discussed in greater detail below, improved techniques are described for
providing continuity check operations to ensure circuit integrity for
communications
over a packet-switched network. The techniques involve the exchange of a known
pattern of bits during call set up processes rather than sending and detecting
tones.
3
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
The techniques can be used to test the integrity of the packet circuit and the
adapters
at either end of the circuit.
The particular examples discussed below involve narrowband traffic such as
voice calls, modem data or facsimile data, sent over a packet-switched
network.
However, continuity checks that include a pattern of bits can be used for
broadband
communications over a packet-switched network as well.
As shown in FIG. 1, a continuous call path can be established starting with a
narrowband Signaling System 7 (SS7) ISDN user part (ISUP) call that
originates, for
example, in a Public Switched Telephone Network (PSTN) 102A. The path can be
established using a virtual circuit over an ATM network 101 and completes on
the
terminating side in a narrowband circuit-switched SS7 ISUP call to the
terminating
subscriber through another circuit switched network 102B. A control mechanism
interacts with the circuit-switched and packet-switched networks to correlate
SS7 and
ATM connections to establish a single continuous information path.
A large number of individual telephone circuits, such as DSO circuits, that
are
to be connected to the packet network 101 can be carried, for example, on
fiber optic
carriers 105 using time-division multiplexing (TDM) according to the Telcordia
Synchronous Optical Network (SONET) standards. The carriers 105 are coupled to
access ports 116 in media gateways 100A, 100B (see FIG. 2).
The media gateways 100A, 100B adapt the TDM telephone line signals to
packet-based signals and vice-versa. The TDM telephone signals are circuit-
switched, in other words, the bit stream can be divided temporally into
individual DSO
circuits. By contrast, in packet-based signals, the bit stream can be divided
according
to the destination address of each packet.
Each gateway 100A, 100B can separate incoming TDM signals into individual
DSO signal streams. In one implementation, shown in FIG. 2, each gateway, such
as
the gateway 100A, includes a TDM switching matrix 117 that provides full
switching
capabilities. The switching matrices 117 permit the DSO circuits to be
interconnected
flexibly with narrowband channels appearing on the gateways.
4
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
Echo cancellation and other digital signal processing functions can be
performed in a digital signal processing portion 118 of each gateway. The
signal
processing portion 118 includes a pattern generator 122 and a pattern detector
124 for
generating and detecting specified patterns of bits, respectively. The pattern
generator
122 and pattern detector 124 can be implemented, for example, using
microprocessors, digital signal processors, or custom application specific
integrated
circuits (ASICs). DSO signal streams are adapted by an ATM adaptation layer
120
into ATM cells. Each cell is inserted through the ATM ports 21 into an ATM
cell
stream 135 that traverses an ATM network 101. The gateways include a control
section 119 that controls overall operation of the gateway. In one
implementation, the
gateways 100A, 100B are implemented as Salix 7720 Class-Independent Switches
available from Tellabs Operations, Inc.
As illustrated in FIG. 1, each gateway 100A, 100B is connected to a respective
ATM end point switch 115. The connection between a gateway and an ATM end
point switch 115 and the connection between the ATM end point switch and the
ATM
network 101 are user-network interfaces (UNIs). Within the ATM network 101,
there
are a number of ATM switches 110 which are interconnected by network-node
interfaces (NNIs).
A call control network 126, which forms part of an existing telephone system,
runs parallel to the voice network. The call control network 126 primarily
controls
telephone switching equipment to connect the originating and terminating ends
of a
telephone call using SS7 messages. A call controller 120A, 120B is coupled to
each
gateway 100A, 100B and provides an interface between the gateway and the call
control network 126. The exchange of call control signals allows the gateways
1 OOA,
100B to establish a connection through the ATM network 101 to enable the
transmission of narrowband traffic between the end points.
As shown in FIGS. 3 and 4, to establish a voice connection, a user at the
originating end dials 210 a telephone number. A connection is established
through an
originating TDM circuit switch in the circuit switched network 102A, and the
call
controller 120A at the originating end receives 215 an SS7 initial address
message
(IAM) 150. The call controller 120A routes 220 the call, in other words, it
identifies a
5
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
call controller 120B associated with a terminating DSO circuit in the circuit
switched
network 102B. The call controller 120A also determines 225 whether a
continuity
check operation is to be performed as part of the call set up. A trade-off
exists
between the desire to perform a continuity check during each call set up and
the extra
time and overhead associated with performing continuity checks. As a result,
typically only a percentage of the call set-ups will include a continuity
check
operation. In one implementation, approximately 5-10% of call set-ups would
include
a continuity check operation.
Assuming that the call controller 120A determines that a continuity check
operation is to be performed, the call controller sends 230 a connection
control
message (CreateConn) 152 to the originating gateway 100A to initiate a
connection
through the ATM network 101. The CreateConn message 152 includes an indication
that a sending-side continuity check operation is requested. In response, the
gateway
I OOA reserves 235 resources for the call and makes the pattern generator 122
and
pattern detector 124 available. Connections are set up between the adaptation
layer
120 and the pattern generator 122 as well as the pattern detector 124. The
pattern
generator 122 repeatedly generates 240 a specified bit pattern, and the
detector 124
monitors 245 the incoming packet stream (if any) for the same bit pattern.
The particular bit pattern used may depend on the application. For example,
in one implementation, the pattern generator 122 is programmed to generate a
bit
pattern such that both binary values in each possible bit position in the
packet can be
checked. Thus, the pattern generator 122 can generate a sequence of two
complementary 8-bit values. Other bit patterns can be generated to allow
various
types of potential failures to be detected and isolated. Thus, in some
implementations,
the pattern generator 122 is programmed to generate a bit pattern comprising a
sequence of 256 values, in other words, a sequence of all possible 8-bit
values. More
generally, the pattern generator 122 can be programmed to generate a sequence
of all
possible n-bit values, where n is the number of bits in each byte.
Another exemplary pattern is illustrated in FIG. 6 and includes twenty 8-bit
values. A first byte includes only "1"s, whereas a second byte includes only
"0"s.
The third through tenth bytes include a single binary "I" with adjacent bytes
differing
6
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
by shifting the binary "1" value from one bit position to an adjacent bit
position.
Similarly, the eleventh through the eighteenth bytes include a single binary
"0" with
adjacent bytes differing by shifting the binary "0" value from one bit
position to an
adjacent bit position. The nineteenth byte includes an alternating pattern
of"1"s and
"0"s. The twentieth byte includes the inverse pattern of the nineteenth byte.
Other bit
patterns may be used depending on the specific errors to be detected. Known
error
detection/correction techniques may be useful in selecting an appropriate bit
pattern
for a given application.
After making the pattern generator 122 and detector 124 available, the
gateway 100A returns 250 an acknowledgement message (CreateAck) 154 that
includes a connection descriptor. The connection descriptor includes an ATM
address
for the gateway 100A as well as information that uniquely identifies the call.
The
information that uniquely identifies the call can identify a connection-
related resource
such as the narrowband circuit (e.g., DSO circuit) handling the call on the
originating
side.
Next, the call controller 120A sends 255 an IAM message 156 to the
terminating call controller 120B. The message 156 includes the information
contained in the connection descriptor as well as an indication that the
continuity
check operation is to be performed. Upon receiving the IAM message 156, the
terminating call controller 120B routes 260 the call. In other words, the
terminating
call controller 120B selects a TDM circuit on a particular gateway, such as
the
gateway 100B, to handle the call. The call controller 120B then sends 265 a
connection control message (CreateConn) 158 to the terminating gateway 100B.
The
CreateConn message 158 also includes the information contained in the
connection
descriptor. In addition, the message 158 includes an indication that the
receiving-side
of a continuity check operation is being requested.
In response, the terminating gateway 100B establishes 270 a packet domain
connection 126 with the originating gateway 100A through the packet network
101,
as shown in FIG. 5. As part of setting up the packet domain connection, the
information that uniquely identifies the call is forwarded through the packet
switches
110, I 15 until it is received by the originating gateway 100A. That allows
the
7
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
originating gateway I OOA to associate the packet-domain connection with the
TDM-
domain connection for the call. The gateways 1 OOA, 1 OOB and ATM switches
110,
115 negotiate the ATM routing headers that will be used between hops along the
packet-domain connection:
The gateway 100B also sets up 275 a continuity check loop between the
incoming and outgoing packet streams 128, 130 associated with the packet
network
connection. FIG. 5 shows the loopback provided in the TDM-domain of the
gateway
100B. More generally, however, the loopback can be provided in either the TDM-
domain or the packet-domain depending on the type of failures the continuity
check is
intended to detect. A control message (CreateAck) 178 is sent 280 by the
terminating
gateway 100B to the terminating call controller 120B to acknowledge that the
packet-
domain connection has been established for the call and that the continuity
check loop
has been set up. The bit pattern appears as successive TDM samples in the TDM
domain, when the loopback is provided in the TDM domain, and appears within
the
I 5 cell stream in the ATM domain.
As the pattern generator 122 repeatedly generates the specified pattern of
bits,
the pattern detector 124 monitors 245 the incoming bit patterns and determines
285
(FIG. 3A) whether the incoming pattern matches the bit pattern that was
generated by
the generator 122. If the packet connection is properly established and if the
adaptation functions in the gateways 100A, I OOB are operating properly, the
specified
bit pattern will be detected by the detector 124 after traversing the packet
connection
116 and the continuity check loop.
The pattern detector 122 can include a software or hardware timer 132 that
provides a timeout function. If the pattern detector 124 does not detect the
generated
pattern within the time set by the timer 132, the continuity check fails. In
that case,
the gateway 100A signals 165 the call controller 120A or a management system
(not
shown) to inform 290 it of the failure. The results of the continuity check
operation
can be used to determine the cause of the failure.
On the other hand, if the generated pattern is detected within the time set by
the timer, then the continuity check is successful. The gateway 100A
disconnects 295
the pattern generator 122 to allow a purge operation to be performed so that
the
8
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
pattern of bits is not forwarded to the TDM circuit handling the call. Once
the
detector 124 no longer detects the pattern, the adaptation layer 116 in the
originating
gateway 100A is connected 300 to the DSO circuit that is handling the call.
If the continuity check is successful, the gateway 100A also notifies 305 its
call controller 120A that the pattern has been detected. The originating call
controller
120A sends 310 an SS7 message to the terminating call controller 120B
informing it
of the successful continuity check. In response, the call controller 120B
instructs the
terminating gateway 100B to disconnect 315 the loopback between the incoming
and
outgoing packet streams. The terminating gateway 100B is then reconfigured 320
to
continue processing the call.
Continuity check packets also can be used as a coarse determination of the
Cell Delay Variation in the packet network. In that case, the pattern of bits
should be
sent over at least several cells.
The foregoing continuity check operations can be used in systems employing
"robbed" bit supervisory signaling as well as clear channel operation.
However, when
the continuity check packets are used in a system employing "robbed" bit
supervisory
signaling, the fact that the low order bits of some frames are used for the
supervisory
signaling should be accounted for. For example, the bits that are used for
supervisory
signaling can simply be ignored for the purpose of the continuity checks.
As described above, different call controllers 120A, 120B are associated with
the gateways 100A, 100B. However, in some cases, both the originating and
terminating gateways 100A, 100B may share a common call controller, such as
the
call controller 120A. In that case, a technique similar to that discussed
above can be
used with a single call controller performing the functions of both call
controllers
120A, 120B. When the call controller 120A routes the call after receiving the
IAM
message 150, it selects the terminating TDM circuit switch and the
corresponding
terminating gateway 100B to handle the call. Also, when a single call
controller
120A is involved, the IAM message 156 need not be used.
9
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
Although the foregoing implementations have been described with respect to
ATM networks, circuit-switched traffic can be routed over other packet-domain
networks, such as frame relay, Ethernet and Internet Protocol (IP) networks,
as well.
Continuity check operations based on a pattern of digital bits are not limited
to
systems under the control of SS7 signaling. In addition, continuity checks can
be
performed independently of call set-up processes. For example, in some cases,
the
pattern generator 122 can continuously generate a pattern over an existing
packet
connection. The pattern detector 124 monitors the return signals and checks
whether
the specified pattern is detected. The output of the pattern detector 124 then
can be
read on demand. Such testing can be used, for example, as part of a
maintenance
program to determine how often failures occur on a particular packet
connection and
its associated gateways.
For situations in which a gateway has multiple adapters for handling
conversions between packet-based and TDM-based bearers, separate pattern
generators 122 and pattern detectors 124 can be provided for each adapter.
Alternatively, the specified pattern can be broadcast over the multiple ATM
channels
and/or connections.
In some implementations, the timer 132 can be incorporated into the call
controller 120A. In that case, the call controller 120A would determine that
the
continuity check had failed if the gateway 1 OOA did not notify it that the
continuity
check was successful within the specified time. Alternatively, the call
controller
120A can be programmed to query the gateway 100A regarding the success of the
continuity check if the gateway has not provided an indication prior to the
specified
time elapsing. The success or failure of the continuity check would then be
determined based on the gateway's response.
Various features of the system can be implemented in hardware, software, or a
combination of hardware and software. For example, some aspects of the system
can
be implemented in computer programs executing on programmable computers. Each
program can be implemented in a high level procedural or object-oriented
programming language to communicate with a computer system. Furthermore, each
such computer program can be stored on a storage medium, such as read-only
CA 02379093 2002-02-06
WO 01/11853 PCT/US00/40571
memory (ROM) readable by a general or special purpose programmable computer,
for
configuring and operating the computer when the storage medium is read by the
computer to perform the functions described above.
Other implementations are within the scope of the claims.
11
