Note: Descriptions are shown in the official language in which they were submitted.
2162185
-- 1 --
FAULT RE~uv~nY 8YSTEN OF A RING NETWORR
BACKGROUND OF THE I~v~llON
This application is a division of co-pending Canadian
Application Serial No. 2,041,789 filed May 3, 1991.
(1) Field of the Invention
The present invention relates to a fault recovery
system of a ring network, and more particularly to a fault
recovery system of a ring network based on a synchronous
transport module (STM) transmission system called a new
synchronization system.
Ring networks based on the synchronous transport
module (STM) transmission system such as a synchronous
digital hierarchy (SDH) or synchronous optical network
(SONET), the standardization of which has been developed in
the CCITT or United States Tl Committee, are expected to be
applied to subscriber systems (Urban Networks) in the
future. The STM transmission system is applied to a high
speed and broad band system of more than 155.52 Mbps. When
a ring network, based on such a STM transmission system
which is a high speed and broad band optical transmission
system, is constructed, the ability of surviving a fault in
the network is important and should be considered from the
beginning of the construction of the system, since a
network fault can have a great influence on the modern
information society.
(2) Description of the Related Art
As a conventionally proposed network fault recovery
system, there are recovery systems employing a loopback
used in a local area network (LAN) and so forth. These
conventional recovery systems, however, are networks based
on packet communication through predetermined protocols,
and therefore, there are problems in that it takes a long
processing time of several seconds to recover from a fault
because the fault must be recovered by the use of the
_ ~1 6~
above-mentioned predetermined protocols. Since the
recovery time in a new synchronization system should be
shorter than, for example, 50 msec, a recovery method which
uses these protocols cannot be employed in a new synchroni-
zation system.
On the other hand, for point to point communication,
the standard usage of automatic protection scheme (APS)
bytes (Kl and K2 bytes in an STM frame) for a switching
control between a working line and a protection line has
been recommended by the CCITT or the United States Tl
Committee. For application to a ring network, however,
standard usage has not been proposed.
A fault recovery system applied to a ring network is
disclosed in Japanese Patent Publication 1-45782, published
on October 4, 1989. This fault recovery system, however,
is not applied to the STM transmission system. Further, in
this document, if multiple faults occur in the working line
and the protection line, the positions of the faults cannot
be determined.
SUMMARY OF THE INVENTION
In accordance with an embodiment of the present
invention there is provided a hybrid type ring network
based on a syn- chronous transport module transmission
system, having a fault recovery system, the hybrid type
ring network comprising: optical fiber transmission lines
including a working line and a protection line running in
opposite directions to each other; and a plurality of
drop/insert nodes connected to each other through the
optical fiber transmission lines, each of the drop insert
nodes including: selecting means for dropping the input
signal from the protection line when the input signal from
the working line is faulty, for dropping the input signal
from the working line when the input signal from the
216~1~5
protection line is faulty and for dropping the input signal
from the working line when both are normal, and passing the
signal as is when the signal is not to be dropped.
In accordance with a preferred embodiment of the
invention each of said drop/insert nodes further includes:
input fault detecting means for detecting an input fault on
the working line or the protection line; fault data writing
means for writing, when the input fault is detected by the
input fault detecting means, fault data in a predetermined
user byte in an overhead of a frame flowing through of the
working line and the protection line; and user type passing
means for passing, when an input fault is not detected by
the input fault detecting means, the user byte as is
through the node.
Preferably, the synchronous transport module
transmission system is a system according to a
recommendation of CCITT G.707, 708 and 709.
It is particularly preferred that the fault data
(Fl(#n,#k,S)) includes a node identification number (#n,#k)
for identifying the node which has detected the input
fault, where S indicates whether the data indicates a fault
report ("O") or a loopback request ("1").
BRIEF DESCRIPTION OF THE DRAWINGS
The above features of the present invention will be
more apparent from the following description of the
preferred embodiments with reference to the accompanying
drawings, wherein:
Figure 1 is a format diagram of an overhead of the STM
frame used in a fault recovery system of a ring network;
Figs. 2A to 2C are principal construction diagrams of
the fault recovery system of a centralized control type
ring network;
Figs. 3A to 3C are principal construction diagrams of
the fault recovery system of a distributed control type
21 62185
ring network;
Figs. 4A to 4D are diagrams showing examples of
various ring constructions used in the present invention;
Fig. 5 is a principal construction diagram of the
fault recovery system of a hybrid ring according to an
embodiment of the present invention;
Figs. 6A to 6C are diagrams for explaining an Fl byte
in the overhead used in the present invention;
Fig. 7 is a block diagram showing a construction
example of a drop/insert node and a supervision node
constructing the centralized control type or the
distributed control type ring;
Figs. 8A to 8C are diagrams showing an example of a
break in the working line causing a fault in the
centralized control type ring;
Figs. 9A to 9C are diagrams showing an example of
breaks in both the working line and the protection line
causing a fault in the centralized control type;
Figs. lOA to lOC are diagrams showing an example of
when plural faults occur in-the centralized control type
ring;
Figs. llA to llC are diagrams showing an example when
both the working line and the protection line are cut
causing a fault in the distributed control type ring;
Figs. 12A to 12C are diagrams showing an example when
plural faults occur in the distributed control type ring;
Fig. 13 is a block diagram showing a construction
example of each drop/insert node in the hybrid ring used in
the system of the present invention; and
Fig. 14 is a diagram showing evaluations of various
fault states in the hybrid ring used in the system of the
present invention.
216218~
DESCRIPTION OF THE PREFERRED EMBODIMENTS
Currently, international standardization of the method
of using the overhead bytes in the STM frame format of the
synchronous transport module (STM) transmission system is
being prepared. In view of this, the inventors of the
present invention have considered using the overhead bytes
for recovering faults in a ring network.
Namely, Fig. 1 shows a frame format of the above-
mentioned STM (in particular the frame format of the STM-
lo 1), in which Al, A2, B1, B2, C1, D1-D12, E1, E2, K1, and
K2, respectively represent bytes, usage of which is already
internationally st~n~rdized. The other F1 byte and Z1 and
Z2 bytes have not yet been internationally standardized but
are determined to be used for domestic or national use.
Note that the remaining bytes are assigned for domestic
use.
Accordingly, by using the unused bytes such as the
latter F1 byte or Z1 and Z2 bytes as user bytes (UB), these
bytes can be utilized to recover a fault in a ring network
of the new synchronization system (hereinafter simply
referred to as a ring network) constructed by optical fiber
transmission lines forming a working line (W) and a
protection line (P) running in opposite directions to each
other.
In the Case of a Centralized Control Type Rinq Network
(Figs. 2A to 2C)
This ring network, as shown in Fig. 2A, is constructed
by drop/insert nodes (nodes A-D for example in the Figure,
and hereinafter also referred to simply as nodes), and a
supervision node SV. Throughout all the embodiments
i
2l62l8~
described in the specification, it is assumed that all of
the nodes are provided in advance with node identification
numbers. First, as shown in the Figure 2A, when an input
fault F (marked by x) of a working line W (or a protection
line P as well) is detected by a node A by detecting a
clock signal error and so forth, the node A writes fault
data FD in a predetermined user byte UB in the overhead of
each of the STM frames flowing through the working line W
and the protection line P, and transmits the STM frames
lo including the written fault data FD to the downstream sides
of the working line W and the protection line P.
When the drop/insert node does not detect an input
fault, it passes the user bytes as they are, and therefore,
to the supervision node SV, the fault data FD in the user
bytes UB are transmitted from the working line W and the
protection line P.
21 6218~
7 --
The supervision node SV determines the position of
the fault F by detecting and analyzing the fault data,
as shown in Fig. 2B, and writes a loopback request LBR,
requiring the nodes A and B located immediately
downstream and upstream of the fault position and
closest to the supervision node SV to execute loopback
operations, into the user byte UB. The written loopback
request LBR is transmitted through the protection line P
and through the node D to the node A. The loopback
request LBR is also transmitted through the working line
W and through the nodes C and D to the node B.
As a result, as shown in Fig~ 2C, the loopbacks are
executed by these nodes A and B to recover the fault.
In this case, instead of utilizing the user byte UB
for transmitting the loopback request, it may be
possible to transmit the loopback request by using K1
and K2 bytes which are internationally standardized. In
this case, it will conform with the international
standardization.
In the Case of a Distributed Type Ring Network (Figs. 3A
to 3C):
In this ring network, there is no supervision node,
and the respective drop/insert nodes are in an equal
relation to each other.
Accordingly, as shown in Fig. 3A, when the
drop/insert node A detects an input fault of the working
line W (or of the protection line P as well), it writes
fault data ED including a loopback request LBR into the
user bytés UB in the STM frames of the working line W
and the protection line P. The user bytes UB including
the fault data FD and the loopback request LRB are
transmitted to the downstream sides of the working line
W and the protection line P.
In a similar h~ay to that shown in Fig. 2A, the
nodes C and D pass the user bytes UB as they are.
Among the respective nodes which received the fault
data, the node B located immediately upstream of the
~ 21 621 85
-- 8 --
input fault and adjacent to the node C which has not
detected an input fault detects that there is a fault
position at the output side of the node B so that, as
shown in Fig. 3B, the node B executes the loopback
operation and returns the loopback request LBR to the
node A which transmitted the loopback requirement LBR.
When the node A receives the loopback request LBR,
the node A executes the loopback operation based on the
returned loopback request LBR so that, as shown in Fig.
3C, the loopback between the nodes A and B is completed
and the fault is restored.
Note, in either of the centralizéd control type or
distributed control type, not only the case where the
ring network is constructed by a working line W and a
protection line P corresponding to each other in a 1:1
relation as shown in Fig. 4A, but a unidirectional ring
constructed by a plurality of working lines and a
single protection line (W1-W3 and P1 in Fig. 4B) is
also possible. A bidirectional ring is also possible
which is constructed by a pair of clockwise and
counterclockwise working lines W1 and W2 and a pair of
counterclockwise and clockwise protection lines P1 and
P2 as shown in Fig. 4C. Further, the fault can also be
restored by constructing a bidirectional ring having a
plurality of pairs of working lines W1 to W6 and a
single pair of protection lines P1 and P2 as shown in
Fig. 4D.
In the Case of a Hibrid-type Ring Network (Fig.5)
In this case also, similar to the distributed
control type shown in Figs. 3A to 3C, there is no
supervision node, and therefore the respective
drop/insert nodes are in an equal relation to each other.
As shown in Fig. 5, when a fault occurs in a point
I between the nodes A and B, the signal from the node A
through the downstream side of the working line W is
transmitted through the nodes D, E, and C until it is
received by the node B, however,the signal from the node
- 21 621 85
g
- A through the protection line P continues to be
transmitted through the loop if there is no
countermesure.
In the hybrid ring, a selecting unit is provided in
each node for selecting a signal to be dropped as
follows. Namely, in each node, if the input signal from
the working line W indicates a fault, the input signal
from the protection line P is dropped; if the input
signal from the protection line P indicates a fault, the
input signal from the working line W is dropped; and if
both input signals are normal, the input signal from
the woxking line W is dropped. If, however, the channel
should not be dropped, it is passed as it is.
And, in this case, similar to the case of Figs. 2A
to 2C and Figs. 3A to 3C, when each node detects an
input fault in the working line W or the protection line
P, fault data is written in the user bytes and is
transmitted downstream of the working line W and the
protection line P, and when an input fault is not
detected, the user bytes are passed as they are, whereby,
in the case of Fig. 5, the fault state between the nodes
A and B can be evaluated with reference to the user
bytes UB.
In the following, embodiments of a fault recovery
system of a ring network relating to the present
invention are described in more detail.
First, as a predetermined user byte in the overhead
used in the system of the present invention, the F1
byte in the STM-1 frame format shown in Fig. 1 is used.
This, however, may be the Z1 or Z2 byte assigned for
the national use, or may be one utilizing various
modifications.
Fig. 6A shows an embodiment of the F1 byte. In this
embodiment, bits bl and b2 are assigned as indicators.
When the bit bl is "O", it indicates whether there is a
fault on the working line; and when it is "1", it
indicates whether there is a fault on the protection
21 621 85
10 --
line, when the bit b2 is "O", it indicates a fault report;
and if the bit b2 is "1", it indicates that the node
identification number of the node which should accept the
loopback request (protection switching) is being conveyed.
Also, the bits b3-b8 are assigned to the identification
number of a node for identifying the node relating to the
fault.
In an F1 byte as above, since only 6 bits can be used
as node number data, when the number of nodes exceeds 26 =
64, it is impossible to identify all nodes using the 6
bits. In this case, two continuous bytes, i.e., a first F1
byte and a second F1 byte of 12 bits as shown in Fig. 6B,
may be utilized. The heading bit bl of the first F1 byte
is defined as "O" so that the first F1 byte conveys fault
data relating to the working line W, and the heading bit bl
of the second byte Fl is defined as "1" so that the second
F1 byte conveys fault data relating to the protection line
P. These bytes are respectively used for detecting faults
on the working line P and the protection line W. An
example of the first F1 byte and the second F1 byte is
shown in Fig. 6C, wherein (1) shows that both the working
line and the protection line are in normal states, (2)
shows that the node "3" on the protection line has detected
an input fault because the bits b7 and b8 in the second F1
byte are "1", and (3) shows that the node "1" on the
working line has detected an input fault because the bit b8
in the first F1 byte is "1".
In the following description, for the sake of
simplicity, the first F1 byte and the second Fl byte are
combined and indicated as Fl (#n,#k, S) where #n indicates
a node number which has detected a fault on the working
line, #k indicates a node number which has detected a fault
in the protection line, and S indicates whether the data
indicates a fault report ("O") or a loopback request ("l'i).
In the following, by using the above-mentioned F1
2l62l8~
byte, the fault recovery system in the above-described
respective rings will be described.
Centralized Control Type Ring
Figure 7 shows an embodiment of a drop/insert node
or a supervision node used in the centralized control
type ring network, which is constructed by a receiving
unit 1 and a transmitting unit 3 for the working line W,
a receiving unit 4 and a transmitting unit 2 for the
protection line P, overhead processing units 5 and 6,
and a data drop/insert/ pass processing unit 7. The
receiving units 1 and 4 are respectively constructed by
light receiving units 11 and 41 connected to the
working line W and the protection line P for converting
light input signals into electrical signals, overhead
dropping units 12 and 42 for dropping the overhead from
the electrical signals to provide to the overhead
processing units 5 and 6, and main signal processing
units 13 and 43 for processing main signals other than
the overheads and for sending the dropped and passed
signals to the data drop/insert/pass processing unit 7.
The transmitting units 2 and 3 are respectively
constructed by main signal processing units 21 and 31
for processing the inserted and passed signals from the
data drop/insert/pass processing unit 7, overhead
inserting units 22 and 32 for inserting overheads from
the overhead processing units 5 and 6 into the inserted
and passed signals, and light transmitting units 23 and
33 for converting the thus generated electric signals
into light signals and for transmitting them to the
protection line P and the working line W, respectively.
Note, the process relating to the overhead is carried
out by the overhead processing units 5 and 6.
~ An Example of a Cut Causing a Fault in the Working
I line W (see Fig. 8)
For the case when a fault is caused by a cut in the
working line W of an optical fiber between the node A
and node B occurs, the fault recovery system of the
21 62I 85
- 12 -
present invention will be described.
Referring to Fig. 8A, the node A, which has
detected, by its light receiving unit 11, the input
fault of the working line W such as a missing clock
signal, loads the node identification number of the
node A on the F1 byte and transmits the F1 byte. In
this case, a fault data Fl(A,-,O) and a loopback
request K(W~ P) are transmitted from the node A to the
downstream side of the working line W through the
communication between the overhead processing units 5
and 6, and fault data Fl(A,-,O) is also transmitted from
the node A to the downstream side of the protection
line P through the communication between the overhead
processing units 5 and 6. The loopback request K(W~ P)
is formed by rewriting the K1 or K2 byte in the STM-1
frame format. Note that, the loopback request K is, as
described with reference to Fig. 1, internationally
standardized. Therefore, when the K1 or K2 byte is used
for the loopback request instead of using the S bit in
the F1 byte, it conforms with the international
standardization. By contrast, when the S bit (b2 in the
F1 byte) is used for the loopback request, the loopback
request can also be executed, and thereforè, the
loopback request K is not always used.
Since the node D is normal, it passes the F1 byte
transmitted from the node A through the working line W,
and the nodes B and C pass the F1 byte transmitted from
the node A through the protction line P. In each of the
nodes D, B, and C, the input signal itself is passed
through the route in which the receiving unit 1, the
data drop/insert/pass processing unit 7, and the
transmitting unit 3 are-connected.
Referring to Fig. 8B, the supervision node SV
detects in its overhead processing units 5 and 6 the
l fault data (F1 byte plus K byte or F1 byte) transmitted
from the node A through the working line W ànd the
protection line P, analyzes the new situation includin~
- _ - 21 621 8.5
- 13 -
the fault data to determine the node A located
imediately downstream of the fault position and closest
to the supervision node SV, and transmits, to the
protection line P, a loopback request (instruction) K(W
~ K) and fault data Fl(A,A,1) requiring-execution of a
loopback at the node A and a fault data Fl(A,A,1). The
supervision node SV also transmits a loopback request
K(W~ P) and a fault data Fl(B,~,1) requiring a loopback
be effected at the node B immediately upstream of the
fault position and closest to the supervision node SV.
The node A detects the loopback request from the
supervision node SV, carries out this loopback, and
then rerturns a loopback response K(W- P) and
Fl(A,A,1) through the working line W to the supervision
node SV. The node B also carries out the loopback, and
then returns a response K(W ~P) and Fl(B,B,1) to the
supervision node SV through the protection line P.
Referring to Fig. 8C, the supervision node SV
receives the loopback responses from the nodes A and B,
and acknowleges that a fault recovery route (loopback
route) has been completed. After completion of the
fault recovery, the supervision node SV resets the F1
byte to be all zeroes and transmits F1(-,-,O) to the
working line W and the protection line P. Accordingly,
in a stationary state where no fault occurs, the
supervision node SV detects the Fl(A,-,O) from the
wor~ing line W and F1(-,B,O) from the protection line P.
~ An Example of Cuts Causing Faults in the Working Line
W and the Protection Line P (see Figs. 9A to 9C)
For the case when ~aultscaused by cuts in both the
working line W and the protection line P between the
node A and the node B occurs, the fault recovery system
of the present invention is described.
I Referring to Fig- 9A, the node A, which has
detected an input fault on the working line W,
transmits Fl(A,-,O) and a loopbac~ request K(W .P) to
the downstream side of the working line W, and transmits
21 621 8~
- 14 -
Fl(A,-,O) to the downstream side of the protection line
P; and the node B, which has detected the input fault
on the protection line P, transmits Fl(-,B,O) to the
downstream of the protection line P.
In this case, even if the Fl~A,-,O) is transmitted
from the node A to the downstream side of the protection
line P, it does not reach the node B because the
optical fiber of the protection line P between the node
A and the node B is cut. Similarly, even when F1(-,B,O)
is transmitted from the node B to the downstrea~ of the
working line W, Fl(-,B,O) on the working line W does not
reach the node A because of the fiber cut ~cut in W).
Since the node D is normal, it passes the F1 byte
transmitted from the node A through the working line W,
and the node C passes the Fl byte transmitted from the
node B through the protection line P.
Referring to Fig. 9B, the supervison node SV
detects the loopback request K~W .P) and the fault data
Fl~A,-,O) transmitted from the node A through the
working line W, detects the fault data Fl~-,B,O)
transmitted from the node B through the protection line
P, analyzes this new situation including the fault data
to determine the position of the ~ault, transmits a
loopback request K(W .P) requiring to loopback at the
node A and Fl~A,A,1) to the downstream side of the
protection line P, and transmits a loopback request K(W
P) requiring a loopback at the node B and Fl~B,B,l) to
the downstream side of the working line W.
The node A detects the loopback request from the
supervision node SV through the protection line P,
carries out this loopback operation, and then returns a
response K~W ,P) and Fl~A,A,1). The node B also
carries out the loopback operation and then returns a
i response K(W P) and Fl(B,B,l) to the supervision node
SV .
Referring to Fig. 9C, the supervision node SV
acknowledges the completion of a fault recovery route
2162185
- 15 -
-
(loopback route) by receiving the loopback responses
from the nodes A and B. After the completion of the
fault recovery, the supervision node Sv resets the F1
bytes to be all zeroes and transmits them to the
working line W and the protection line P. Accordingly,
in the stationary state in which there is no fault in
the ring network, the supervision node SV detects the
Fl(A,-,O) from the working line W and the F1(-,B,O)
from the protection line P.
~ An Example of Plural Faults (see Figs. lOA to lOC)
For the case when both the working line W and the
protection line P between the nodes A and B are cut and
when the protection line P is cut between the nodes D
and A, the fault recovery system of the present
invention will be described.
Referring to Fig. lOA, similar to the above example
shown in Fig. 9A, the node A transmits a loopback
request K(W ~P) and Fl(A,A,O) through the downstream
side of the working line W to the supervision node SV,
and the node B transmits fault data F1(-,B,O) through
the downstream side of the protection line P to the
supervision node SV.
Referring to Fig. lOB, the supervision node SV
analyzes the new situation including the fault state to
determine the position of the fault. Then the
supervision node SV transmits a loopback request K(W ~P)
requiring a loopback at the node D and El(D,D,1) to the
downstream side of the protection line P, and transmits
a loopback request K(W ~P) requiring to loopback at
the node B and Fl(B,B,1) to the downstream side of the
working line W. Then, the node D detects the loopback
request from the supervision node SV, carries out this
loopback operation, and returns a response K(W~ P) and
I Fl(D,D,1) to the supervision node SV. The node B also
detects the loopback request from the supervision node
SV, carries out this loopback operation, and returns a
response K(W .P) and Fl(B,B,1) to the supervision node
21 62I 8~
- 16 -
SV .
Referring to Fig. lOC, the supervision node SV
ac~nowledges the completion of a fault recovery route
(loopback route) by receiving the loopback responses
from the nodes D and B. After the completion of the
fault recovery, the supervision node SV resets the F1
bytes to be all zeroes and transmits them to the
working line W and the protection line P. Accordingly,
in a stationary state in which there is no fault, the
supervision node SV detects Fl(D,-,O) from the working
line W and F1(-,B,O) from the protection line P.
Distributed Control Type Ring
In this distributed control type ring network,
there is no supervision node, and the respective
drop/insert nodes are placed in an equal relation to
each other. In this case also, the example of the
construction shown in Fig. 7 can be applied, and the
difference from the centralized control type ring is
that, since there is no supervision node, the F1 bytes
are not reset by the supervision node, and the others
are processed in a similar way to those in the
centralized control type ring.
~ An Example of Cuts Causing Faults in the Working Line
W and the Protection Line P (see Figs. llA to llC)
For the case when both the working line W and the
protection line P between the nodes A and B are cut,
the fault recovery system of the present invention will
be described.
Referring to Fig. llA, the node A which has
detected a fault on the working line W transmits Fl(A,*,
O) and a loopback request K(W~ P) to the downstream
side of the working line W, and transmits Fl(A,*,O) to
the downstream side of the protection line P. In this
case, in the initial state of the fault, there is a
possibility that the node A may not know about the fault
on the protection line P because the fault on the
protection line P is at the output side of the node A,
~ 1 b~ 1~5
- 17 -
so that the node A may transmit Fl(A,*,O) to the
downstreams of the working line W and the protection
line P. The node A, however, will know about the fault
on the protection line P from the later received fault
data from the.node B thorugh the protection line P.
After the node A acknowleges the fault on the protection
line P, the node A transmits Fl(A,B,-) to the
downstream sides of the working line W and the
protection line P. In this case,~ represents a time
dependent parameter. Note, in this case also, if the K
bytes of the loopback request are not used as mentioned
before, a loopback request to an other node is realized
by changing the "O" in the Fl(A,~ , O) to "1".
The node B which has detected a fault on the
protection line P transmits Fl(*,B,O) of the F1 bytes
cont~; ni ng the node identification number of the node B,
to the downstream sides of both the working line W and
the protection line P. In this case, the Fl(A,*,O) on
the protection line P does not reach the node B because
of the cut fiber (P cut), and the F1(~ ,B,O) on the
working line W does not reach the node A because the
cut fiber (W cut). The nodes D, E, and C pass the F1
byte transmitted from the node A through the working
line W, and the nodes C, E, and D pass the F1 byte
transmitted from the node B through the protection line
P.
Referring to Fig. llB, the node B detects the
loopback request K(W - P) and the fault data including
the Fl(A,*,O) transmitted from the node A through the
working line W. By analyzing the fault data Fl(A,*,O),
the node B recognizes that its own node B is placed
immediately upstream of he fault position, i.e., just
before the node A which has transmitted the fault data.
This recognition is possible because all-of the nodes
are provided with their own node identification numbers
in advance. Accordingly, the fault on the protection
line P between the nodes A and B is determined so that a
2162185
- - 18 -
-
loopback operation is executed at the node B. The node
B then transmits a loopback response K(W .P) and
Fl(A,B,O) through the protection line P to the node A.
Referring to Fig. llC, the node A receives the
loopback response K(W .P) and the Fl(A,B,O) from the
node B through the protection line P so that it
determines the position of the fault on the working line
W between the nodes A and B. Then the node A executes a
loopback operation from the protection line P to the
working line W. In this way, it is recognized that the
fault recovery route (loopback route) has been
completed, and, in a stationary state after the fault
recovery completion, the Fl(A,B,O) is transmitted
through the working line W and the protection line P.
An Example of Plural Faults (see Figs. 12A to 12C)
For the case when faults occur due to cut in both
the working line W and the protection line P between
the node A and the node B, and a cut in the working line
W between the nodes B and C, the fault recovery system
of the present inveniton will be describèd.
Referring to Fig. 12A, the node A transmits a
loopback request K(W~ P) and Fl(A,*,O) through the
downstream side of the working line W to the node C,
and the node ~ transmits Fl(B,B,O) through the
downstream side of the protection line P to the node A
since both the working line W and the protection line P
are in their input fault states.
Referring to Fig. 12B, the node C detects the fault
on the working line W between the node C and the node B
based on the fault data Fl(B,B,O) transmitted from the
node B through the protection line P and the fault data
Fl(A,~ ,O) from the node-A. The node C analyzes this
new situation, executes a loopback operation, and
i transmits a loopback response k(W~ P) and Fl(A,C,O) to
the node A throu~h the protection line P.
Referring to Fig. 12C, the node A, which recei~es
this loopback response K(W ~P) and Fl(A,C,O) from the
- 19 - 2l62l85
node C, determines the fault on the working line W
between the nodes A and B. Then, the node A executes
the loopback operation at the node A to complete the
fault recovery route (loopback route). In the
stationary state after completion of the fault recovery
completion, Fl(A,C,O) is transmitted through the
working line W and the protection line P.
Hybrid Ring
In this case also, there is no supervision node so
that the respective nodes have an equal relation to each
other.
Figure 13 schematically shows the construction of
each node. In the figure, the same references as those
in Fig. 7 represent the same parts, and, also included
are a selector 8 for selecting data to be dropped or
passed from or the receiving unit 1 or 4 to the data
drop/insert/pass processing unit 7, a distributing unit
9 for distributing the inserted or passed data from the
data drop/insert/pass processing unit 7 to the
transmitting units 2 or 3, and a control circuit 10 for
controlling the selector 8 . The control circuit 10
selects a normal signal from the signals received by the
receiving units 1 and 4. When both are normal, the
control circuit 10 selects the receiving signal on the
working line W. However, the selector 8 is controlled
only when the drop/insert mode is effected for the
corresponsing channel chich needs to be dropped or
inserted. In the case of another channel, namely, when
the channel is not to be dropped or inserted, the
receiving units 1 and 4 and the transmitting units 3 and
2 are connected straight through as illustrated in the
figure by a dotted line. Note, the discrimination of
whether or not the signal is normal can be carried out
based on a cut in the input signal or by a lack of frame
synchronization. It may also be discriminated by an
alarm indication or a pointer indicating an abnormal in
H1 and H2 pointer bytes included in the overhead of the
- 20 -
STM frame which is processed by the overhead processing
units 5 and 6.
Examples of faults in the hybrid ring using the
no~e having such a construction as described above, are
illustrated in Fig. 14.
(1) An Example When the Wor~ing Line W Between the Nodes
A and B is Cut
In this case, as the F1 byte of the fault data, the
F1 byte Fl(A,-,0) which is the same as the one shown in
Figs. 8A to 8C is output from the nodes A and B (time
tl). In the node A, since an input fault occurs on the
wor~ing line W, only the received signal transmitted
from the node D through the protection line P is
determined as normal and is received byte by byte. Also,
in the node B, the received signal from the node C
through the working line W and the received signal from
the node A through the protection line P are received as
normal and are received byte by byte. Therefore, the
control circuit 10 in the node B switches the selector 8
to receive with priority the received signal from the
working line W. Note, the other nodes C, D, and E only
pass the received signals on the working line W and the
protection line P.
After this, even at a time t2 when some time has
passed from the time tl, the state of the F1 byte is
quite unchanged.
Thus, by the wor~ing line W and the protection line
P, the node A and the node B communicate with each
other without loopback.
Also, since the overhead is used in this case also,
the fault evaluation (of a cut in the working line W
between the nodes A and B) can be executed in a way
similar to that mentioned above in the nodes A and B.
(Z) An Example When Both the Wor~ing Line W and the
Protection Line P Between the Nodes A and B are cut
In this case, the F1 bytes Fl(A,-,0) and F1(-,B,0)
as the fault data, which are the same F1 byte as shown
2162I8~
- 21 -
in Figs. 9A to 9C and Figs. llA to llC, are output from
the nodes A and B, respectively (time tl). The node A
receives only the receiving signal transmitted from the
node D through the protection line P as a normal signal
because the input fault occurs on the working line W
between the nodes A and B, and the node B receives only
the receiving signal transmitted from the node C through
the working line W as a normal signal because the input
fault occurs on the protection line P between the nodes
A and B.
After this, at a time t2, since both the node A and
the node B respectively know about the the faults on
the protection line P and the working line W, the F1
bytes become Fl(A,B,O) as illustrated in Fig. 14 (2).
Thus, by the working line W and the protection line
P, the node A and the node B communicate with each
other without loopback.
Also, by the use of the usdbyte in this case also,
the fault evaluation (of a cut in the working line W
between the nodes A and B) can be executed in the nodes
A and B in a way similar to that mentioned before.
(3) An Example When Both the Working Line W and the
Protection Line P Between the Nodes A and B are Cut and
the Working line W between the Nodes B and C is Cut
In this case, the F1 bytes Fl(A,-,O) and Fl(B,B,O)
as the fault data, which are the same F1 bytes as shown
in Figs. lOA to lOC and Figs. 12A to 12C, flow through
the working line W and the srotection line P (time tl).
The node A receives only the receiving signal
transmitted from the node D through the protection line
P as a normal signal because the input fault occurs on
the working line W between the nodes A and B, the node B
cannot receive a signal because the input faults occur
on both the working line W and the protection line P,
and the node C receives with priority the receiving
signal transmitted from the node E through the working
line W as a normal signal.
216218~
- 22 -
At a time t2 after a certain amount of time has
passed, the node A detects the input fault of the node
B so that the F1 byte Fl(A,B,O) as illustrated in the
figure flows through the working line W.
Thus, through the working line W and the protection
line P, the nodes A and C communicate with each other
without loopback.
In this case also, since the overhead is used, the
fault evaluation (of a cut in the working and protection
lines between the nodes A and B, and a cut in the
working line W between the nodes B and C) can be
executed in the nodes A, B, and C in a way similar to
that mentioned before.
Thus, in the hybrid ring also, by applying the
overhead, the ability to respond to a fault in a ring
(especially the case of a plurality of faults or a
catastrophic fault) can be increased.
As described above, according to the fault
recovery system of a ring network relating to the
present invention, by utilizing a predetermined user
byte in the overhead of the ST~ frame used in the
synchronous transport module transmitting system, an
input fault detected in any node in a centralized
control type ring, distributed control type ring, or
hybrid ring is transferred to an other node, whereby the
supervision node or the drop/insert node detects the
position of the fault to execute a loopback operation or
a hybrid process. Therefore, since no protocol is used
in the fault recovery process, the fault can be
recovered in a short time.
