Note: Descriptions are shown in the official language in which they were submitted.
~z79~32
RING TRANSMISSION SYSTEM
This invention relates to a transmission system comprising nodes,
for example telephone central offices, coupled in a ring via two
oppositely directed transmission paths.
Such transmission systems are generally known and described in
various references. For example, Smith et al. U.S. Patent No. 3,859,468
issued January 7, 1975 and entitled "Redundant Data Transmission
Arrangement" describes an arrangement in which transmission terminals are
serially connected to a base terminal by primary and secondary looped
lines over which tdm (time division multiplexed) data signals are
transmitted in opposite directions. Seo U.S. Patent No. 4,446,551 issued
May 1, 1984 and entitled "Data Highway System with Dual Transmitting Loop
Lines", and Nakayashiki et al. U.S. Patent No. 4,506,357 issued March 19,
1985 and entitled "Method and Apparatus for Switching Loop Type
Transmission Lines", describe similar types of transmission system in
which a master station and remote or slave stations are coupled in a
ring via oppositely directed transmission paths. Jones et al. U.S.
Patent No. 4,633,246 issued December 30, 1986 and entitled "Time Division
Multiplex Ring" describes a ring transmission system with oppositely
directed transmission paths coupling nodes which are all equivalent to
one another; i.e. there is no master station.
In such known systems, one of the transmission paths is a main or
primary transmission path which is used in normal operation, each node
receiving signals from and transmitting signals to this path in a single
direction around the ring. The other transmiss;on path is a standby or
secondary path which is used to protect transmitted signals in the event
of various fault conditions occurring.
For example, in the Jones et al. patent, in normal operation each
node transmits signals to and receives signals from the main path which
has a clockwise transmission direction around the ring. The same signals
are transmitted to, but are not received from, the standby path which has
a counter-clockwise transmission direction. In the event of a fault
occurring, the nodes may receive signals from the standby path instead of
from the main path, may bridge the standby path through the node, or may
loop back signals from the main path to the standby path, to maintain
communications among the nodes. In a fault condition, the two loops are
"` lZ79~.3~
reconfigured to form a single folded loop. Various fault conditions,
and the resulting reconfiguration, are described in the Jones et al.
patent.
An object of this invention is to provide an improved ring
transmission system, and an improved method of transmitting signals in
such a system.
According to one aspect this invention provides a transmission
system comprising a plurality of nodes coupled in a ring via first and
second multiplexed transmission paths providing for transmission in
opposite directions around the ring, the nodes comprising means for
providing, in normal operation, communications from a first node to a
second node via the first path and from the second node to the first
node via the second path, and for providing, in the presence of a fault
affecting said communications in normal operation, communications from
the first node to the second node via the second path and from the
second node to the first node via the first path, whereby bidirectional
communications between the first and second nodes are provided in
normal operation via both paths on a first part of the ring between the
nodes and in the presence of the fault via both paths on a second,
different, part of the ring between the nodes; wherein there is a third
node is coupled in the first part of the ring between the first node
and the fault, and the means for providing communications in the
presence of the fault comprises means for providing communications:
from the first node to the second node via the first path between the
first and third nodes, via the third node between the first and second
paths, and via the second path between the third and second nodes; and
from the second node to the first node via the first path between the
second and third nodes, via the third node between the first and second
paths, and via the second path between the third and first nodes.
According to another aspect this invention provides a
transmission system comprising a plurality of nodes coupled in a ring
via first and second multiplexed transmission paths providing for
transmission in opposite directions around the ring, the nodes
comprising means for providing, in normal operation, communications
from a first node to a second node via the first path and from the
second node to the first node via the second path, and for providing,
in the presence of a fault affecting said communications in normal
~.~7~313~
operation, communications from the first node to the second node via
the second path and from the second node to the first node via the
first path, whereby bidirectional communications between the first and
second nodes are provided in normal operation via both paths on a first
part of the ring between the nodes and in the presence of the fault via
both paths on a second, different, part of the ring between the nodes;
wherein a third node is coupled in the first part of the ring
between the first node and the fault and a fourth node is coupled in
the first part of the ring between the second node and the fault, and
the means for providing communications in the presence of the fault
comprises means for providing communications: from the first node to
the second node via the first path between the first and third nodes,
via the third node between the first and second paths, via the second
path between the third and fourth nodes, via the fourth node between
the second and first paths, and via the first path between the fourth
and second nodes; and from the second node to the first node via the
second path between the second and fourth nodes, via the fourth node
between the second and first paths, via the first path between the
fourth and third nodes, via the third node between the first and second
paths, and via the second path between the third and first nodes.
The invent;on provides the advantage that nodes which are not
adjacent to a fault need not be aware of the fault and do not need to
take any protection switching action in response to the fault. In
other words, in the event of a fault anywhere around the ring, only the
nodes on each side of and immediately adjacent to the fault need to be
aware of the fault to take protection switching action.
The invention also extends to a communications network
compr;sing a plurality of transmission systems, each as recited above,
intersecting at at least one network node, the network node comprising
a node of each transmission system and switching means for providing
communications between the transmission systems.
Another aspect of the invention provides a method of
transmitting signals in a transmission system comprising at least four
nodes coupled in a ring via first and second transmission paths
providing for transmission in opposite directions around the ring,
comprising the steps of: in normal operation, transmitting signals
from a first node to a second node via the first transmission path and
~: .
~7~ 32
a third node between the first and second nodes, and transmitting
signals from the second node to the first node via the second
transmission path and the third node; and in the presence of a fault
between the second and third nodes, transmitting signals from the first
node to the second node via the first transmission path between the
first and third nodes, via the third node between the first and second
transmission paths, and via the second transmission path and at least a
fourth node between the third and second nodes, and transmitting
signals from the second node to the first node via the first
transmission path and the at least fourth node between the second and
third nodes, via the third node between the first and second
transmission paths, and via the second transmission path between the
third and first nodes.
The invention will be further understood from the following
description with reference to the accompanying drawings, in which:
Figs. 1 and 2 illustrate a known ring transmission system in
normal and fault conditions respectively;
Figs. 3 and 4 illustrate a ring transmission system in
accordance with an embodiment of this invention in normal and fault
conditions respectively;
Fig. 5 illustrates a node of the system of Figs. 3 and 4;
Fig. 6 illustrates a transmission system in accordance with
another embodiment of this invention in a fault condition; and
Fig. 7 illustrates a node of the system of Fig. 6; and
Fig. 8 illustrates a cross connect node in a network of
intersecting loops.
Although for clarity and simplicity Figs. 1 to 4 each show a
transmission system with four nodes, and Fig. 6 shows a system with six
nodes, in each case the transmission system can have an arbitrary
number of nodes.
Referring to Figs. 1 and 2, a known ring transmission system is
illustrated with four nodes coupled in series via main and standby
transmission paths 10 and 12 respectively. Each of the paths 10 and 12
is in the form of a ring or loop for carrying a signal in a respective
direction around the loop, clockwise for the main path 10 and counter-
clockwise for the standby path 12. Each node includes a transmit port
~ Z''~9 ~L3~2
T and a receive port R, at which arbitrary channels of the signals
carried around the loop can be dropped or added respectively.
Fig. 1 illustrates the system in normal, fault-free, operation.
In each node signal channels incoming on the main path 10 are either
dropped at the transmit port T, and replaced by signal channels added
at the receive port R, or are passed through to the outgoing port on
the main path 10. The signal outgoing on the main path 10 is also
supplied to the outgoing port on the standby path 12, but there is no
connection for signal channels incoming on the standby path 12.
Thus for example for a signal channel connection between the
nodes 1 and 2, the signal incoming at the receive port R of node 1 is
coupled via the main path 10 to the transmit port T of node 2, and the
signal incoming at the receive port R of node 2 is coupled via the main
path 10, passing through the nodes 3 and 4 in turn, to the transmit
port T of node 1.
Fig. 2 illustrates the known system in a fault condition in
which both of the main and standby paths 10 and 12 are interrupted
(e.g. due to a cable cut) between the nodes 1 and 2. In this
situation, within the node which on the main path 10 is downstream from
the interruption, i.e. within node 2 in this case, there is no
connection for any signal incoming on the main path 10, and instead the
signal incoming on the standby path 12 is connected and is looped back
to the outgoing port on the main path 10. Within each node which is
not adjacent to the interruption, i.e. within each of nodes 3 and 4 in
this case, the standby path 12 is bridged through the node. In other
words, the outgoing port on the standby path is connected to the
incoming port on the standby path instead of to the outgoing port on
the main path.
The reconfigured system in Fig. 2 provides a single folded loop
for communication among all the nodes. For example, for a signal
channel connection between the nodes 1 and 2, the signal incoming at
the receive port R of node 1 is coupled via the standby path 12 and the
bridged through connections in the nodes 4 and 3 in turn to the
transmit port T of node 2, and the signal incoming at the receive port
R of node 2 is coupled via the main path 10 and the nodes 3 and 4 in
turn to the transmit port T of node 1. Similarly, for a signal channel
connection between the nodes 3 and 4, a signal at node 3 port R is
I
79132
coupled via the main path 10 to node 4 port T, and a signal at node 4
port R is coupled via the main path 10 to node 1, then via the standby
path 12 and nodes 4 and 3 in turn to node 2, then via the loop-back
connection to the main path 10 and thence to node 3 port T.
From the above description, it can be seen that in the system of
Figs. 1 and 2 all bidirectional traffic between any two nodes is
carried once around a loop, this being constituted by the main path 10
in the normal situation of Fig. 1 and being constituted by the folded
loop formed by both paths 10 and 12 in the fault condition of Fig. 2.
10 Thus each of the paths 10 and 12 must have the capacity between any two
nodes to carry the total of all signal channels in the loop.
By way of example, suppose that there is only point-to-point
traffic between adjacent nodes (no point-to-point traffic between non-
adjacent nodes) of 11 signal channels between nodes 1 and 2, 5 signal
channels between nodes 2 and 3, 24 signal channels between nodes 3 and
4, and 8 signal channels between nodes 4 and 1. Then each of the main
and standby paths 10 and 12 must be capable of carrying the sum (11 + 5
+ 24 + 8) of 48 signal channels. It should be appreciated that the
numbers of signal channels carried between nodes of the system can
differ in this manner around the loop because traffic is dropped and
added at each node. This dropped and added traffic may include traffic
circulating in other, adjacent and overlapping, loops of a more complex
network of which the system of Figs. 1 and 2 may form a part7 as is
explained further below with reference to Fig. 8.
Figs. 3 and 4 illustrate a ring transmission system in
accordance with an embodiment of the invention, using where appropriate
similar terminology to that used above with reference to Figs. 1 and 2.
Fig. 3 illustrates normal, fault-free, operation, and Fig. 4
illustrates operation in a fault condition similar to that of Fig. 2.
In the system of Figs. 3 and 4 there are four nodes and two paths, for
clockwise and counter-clockwise sisnal transmission, as in the system
of Figs. 1 and 2, but as described further below the paths do not
constitute a main and a standby path as in the prior art, and are
accordingly referred to as a first path 14 for clockwise transmission
and a second path 16 for counter-clockwise transmission.
Referring to Fig. 3, showing normal, fault free, operation, in
each node signal channels incoming on the first path 14 are selectively
127~?13Z
7 `
coupled to the node's transmit port T or passed through to the outgoing
port on this path 14, signal channels incoming on the second path 16
are selectively coupled to the node's transmit port T or passed through
to the outgoing port on this path 16, and signal channels incoming at
the node's receive port R are selectively supplied to either the
outgoing port on the path 14 or the outgoing port on the path 16.
In this case, for a signal channel connection between the nodes
1 and 2, the signal incoming at the receive port R of node 1 is coupled
via the first path 14 to the transmit port T of node 2, and the signal
incoming at the receive port R of node 2 is coupled via the second path
16 to the transmit port T of node 1. Thus for such a signal channel
connection between adjacent nodes, signals are coupled via both of the
paths 14 and 16 directly between the nodes in the manner of
bidirectional point-to-point communications, and the remainder of the
loops on the paths 14 and 16 do not carry these s;gnals. More
generally, in the system of Fig. 3 signals between any two nodes are
coupled via the most direct route around the loop on both paths for the
two directions of transmission, the node coupling the signals between
its ports to this end. This can be seen to be quite different from the
system of Fig. 1, in which only one of the paths is effective in
normal, fault-free, operation.
Fig. 4 illustrates the same system as Fig. 3 but in a fault
condition which is the same as that in Fig. 2, i.e. an interruption of
both paths 14 and 16 between the nodes 1 and 2. In Fig. 4, the nodes 1
and 2 do not couple signals to, or receive signals from, the faulty
parts of the paths. Thus now the node 1 couples its receive port R
only to the outgoing port on the path 16 and couples its transmit port
T or,ly to the incoming port on the path 14, and conversely the node 2
couples its receive port R only to the outgoing port on the path 14 and
couples its transmit port T only to the incoming port on the path 16.
Neither of the nodes 1 and 2 prov;des pass-through coupl;ng on the
paths 14 and 16. The connection paths between the ports of the nodes 3
and 4 not on either side of the fault are unchanged from those shown in
Fig. 3 (although the signal channel couplings between the ports of
these nodes may be changed as described further below).
The reconFigured system in Fig. 4 also maintains bidirectional
communications among all of the nodes. For example, for a s;gnal
9~132
channel connection between the nodes 1 and 2, the signal incoming at
the receive port R of node 1 is coupled via the path 16 and the pass-
through connections in the nodes 4 and 3 in turn to the transmit port T
of node 2, and the signal incoming at the receive port R of node 2 is
coupled via the path 14 and the pass-through connections in the nodes 3
and 4 in turn to the transmit port of node 1. Thus bidirectional
point-to-point communications between the nodes 1 and 2 are maintained,
but now the coupling of signals is via the less direct, still intact,
part of the loop on each path.
In the system of Figs. 3 and 4, a signal channel connection
between, for example, the nodes 3 and 4 is unaffected by the fault
between the nodes 1 and 2; in both the normal and the fault condition,
signals are coupled from node 3 port R via the path 14 to node 4 port
T, and from node 4 port R via the path 16 to node 3 port T. Similarly,
connections between the adjacent nodes 1 and 4, and between the
adjacent nodes 2 and 3, and connections between the nodes 1 and 3 via
the node 4, and between the nodes 2 and 4 via the node 3, are
unaffected by the fault between the nodes 1 and 2.
In the system of Figs. 3 and 4, a signal channel connection
between the nodes 1 and 3 v;a the node 2 is affected by the fault
between the nodes 1 and 2. In this case, the signal incoming at node 3
port R, and previously coupled via the path 16 and node 2 to node 1
port T, must in Fig. 4 instead be coupled via the path 14 and node 4.
Conversely, the signal supplied to node 3 port T, originating at node 1
port R and previously derived via node 2 and the path 14, must in Fig.
4 be derived via node 4 and the path 16. Accordingly, the node 3
remote from the fault must nevertheless be informed (through signalling
between the nodes on overhead information) of the fault so that it can
effect appropriate switching of signal channel connections which are
affected by the fault and are terminated (i.e. are supplied to and from
the ports T and R respectively) at this node. Similar comments apply
in respect of the node 4.
Using the same numbers as before, suppose that there is only
point-to-point traffic between adjacent nodes (no point-to-point
traffic between non-adjacent nodes) of 11, 5, 24, and 8 signal channels
between nodes 1 and 2, 2 and 3, 3 and 4, and 4 and 1 respectively.
Then in the fault condition of Fig. 4, the paths 14 and 16 must be able
~Z~132
to carry 5 + 11 = 16 signal channels (i.e. the original 5 channels
between nodes 2 and 3 plus the redirected 11 channels from between
nodes 1 and 2) between nodes 2 and 3, 24 + 11 = 33 signal channels
between nodes 3 and 4, and 8 + 11 = 19 signal channels between nodes 4
and 1. In general, in this simple case of only adjacent node traffic,
for protection against interruption of both paths 14 and 16 anywhere
around the loop, the two paths 14 and 16 between any two adjacent nodes
must be capable of carrying the normal point-to-point traffic between
these two nodes plus, for protection, the greatest level of point-to-
point traffic between any other two adjacent nodes around the loop.This is obviously no greater than, and is normally less than, the sum
of all traffic anywhere in the loop as in the prior art system of Figs.
1 and 2.
Thus in the system of Figs. 3 and 4 the capacity required of the
paths 14 and 16 for carrying traffic in both normal and fault
conditions is generally less than that required of the paths 10 and 12
in the prior art system of Figs. 1 and 2. This saving in capacity
means that the paths 14 and 16 can have a lower capacity than the paths
10 and 12 for the same traffic conditions, and/or the paths 14 and 16
can have the same capacity as the paths 10 and 12 while providing a
greater traffic capacity between the nodes.
As should be appreciated from the above description, the
capacity saving arises from the switching within the nodes of traffic
to both paths 14 and 16, whereby both paths are used for bidirectional
traffic between any two nodes via the most direct path available. In
consequence, each individual path 14 or 16 carries both its normal
traffic and protected (;.e. redirected) traffic in the fault condition
of Fig. 4, and in the normal condition carries its normal traffic and
has spare capacity for protected traffic in the event of a fault.
In a normal system traffic will be present not only between
adjacent nodes (e.g. nodes 1 and 2) as discussed above, but also
between non-adjacent nodes (e.g. nodes 1 and 3). The system of Figs. 3
and 4 does not provide a capacity saving, compared with the system of
Figs. 1 and 2, for traffic between non-adjacent nodes, because such
traffic travels half way around the loop in each case, but there will
still be the saving as discussed above for the traffic between adjacent
nodes. In general, for loops with arbitrary numbers of nodes, the
79132
invention provides a capacity saving except in respect of traffic which
travels half way around the loop.
Fig. 5 illustrates a node of the system of Figs. 3 and 4 for use
in a synchronous optical network operating in accordance with the
proposed SONET standard, referred to below simply as SONET. Before
describing Fig. 5, it is appropriate to describe SONET briefly.
In SONET, a basic format is provided for a signal at a bit rate
of 51.84Mb/s. Such a signal in electr;cal form is referred to as an
STS-1 signal, STS being an abbreviation for Synchronous Transport
Signal. A plurality of n such signals may be byte interleaved to form
higher level synchronous signals, referred to as STS-n signals. Thus
an STS-3 signal (n = 3) has a bit rate of 3 x 51.84 = 155.52Mb/s, and
an STS-12 signal (n = 12) has a bit rate of 12 x 51.84 = 622.08Mb/s.
A conventional DS-3 signal at a bit rate of 44.736Mb/s, or 28
conventional DS-1 signals, each within a so-called virtual tributary at
a bit rate of 1.728Mb/s, can be mapped into an STS-1 signal to be
carried via a SONET network. Similarly, higher level existing signals
at bit rates of the order of 140Mb/s and 570Mb/s can be mapped into
STS-3 and STS-12 signals respectively.
An optical carrier signal, which is generated by converting an
STS-n signal into an optical signal to be transmitted via an optical
fiber path, is referred to as an OC-n signal. OC-12 signals at a
modulated bit rate of 622.08Mb/s, and OC-48 signals at a modulated bit
rate of 2.48832Gb/s, are of particular significance in SONET.
The synchronous nature of the STS-n signals makes it relatively
simple to access, for example, individual STS-1 signals within an STS-
12 signal, whereby overhead information in each STS-1 signal can be
readily separated for monitoring, control, and signalling purposes.
This overhead information includes automatic protection switching (APS)
bytes, designated K1 and K2, for each STS-1 signal, which are used for
communication from the tail end to the head end of a connection to
effect protection switching for the STS-1 signal.
As the SONET format is byte interleaved, with 8-bit bytes,
within each node the signals are conveniently handled in 8-bit parallel
form. Thus an STS-12 signal has a bit rate of 622.08/8 = 77.76Mb/s on
each of eight parallel lines.
~Z'79~3Z
11
Referring now to Fig. 5, the incoming and outgoing ports of the
path 14 are referenced 14' and 14" respectively, and the incoming and
outgoing ports of the path 16 are referenced 16' and 16" respectively.
Each of these ports is coupled to a respective one of four electro-
optical interfaces (E/0) 18, 20, 22, and 24 respectively. The node'stransmit port T and receive port R are coupled to an interface circuit
26.
Between the output of the E/0 18 and the input of the E/0 20 are
connected in series a serial-to-parallel converter and overhead
demultiplexer circuit (S/P & 0/H) 28, a time slot interchanger (TSI)
30, an STS synchronizing circuit (SYNC) 32, another TSI 34, a selector
(SEL) 36, and an overhead multiplexer circuit and parallel-to-serial
converter (0/H ~ P/S) 38. Similarly, the output of the E/0 22 is
connected via an S/P & 0/H 40, a TSI 42, a SYNC 44, a TSI 46, an SEL
48, and an 0/H & P/S 50 to the input of the E/0 24. The node's receive
port R is coupled via the interface circuit 26 and two further TSIs 52
and 54 to second inputs of the selectors 36 and 48 respectively, and
outputs of the SYNCs 32 and 44 are connected via a selector (SEL) 56
within the interface circuit 26 and thence to the node's transmit port
20 T. In addition to the selector 56, the interface circuit 26 provides
the necessary interfacing (including multiplexing, demultiplexing,
synchronization, etc. between the STS signals, supplied by the SYNCs 32
and 44 and supplied in the TSIs 52 and 54, and the signals at the ports
T and R, whether these are in an STS form as assumed herein or another
(e.g. DS-3) form.
The node further includes a control processor block 58 which is
connected (via lines not shown) to all of the elements 26 to 56 for
deriving information therefrom and for the control thereof. Functions
of this block 58, which includes connection and selection memories for
the TSIs and selectors and also serves for maintenance purposes, will
become apparent from the following description. In addition, within
the node various of the elements shown in Fig. 5 and the connections
therebetween may be duplicated in known manner for redundancy and
reliability, but for clarity and simplicity the duplicated equipment
is not described further here.
It is assumed here that each of the paths 14 and 16 carries an
OC-48 optical signal, so that such a signal is present at each of the
-- ~Z7913;~
12
incoming ports 14' and 16'. These signals for the two directions of
transmission through the node are processed similarly, and accordingly
the following description relates only to signals derived from the port
14'.
The OC-48 signal at the port 14' is converted into an STS-48
electrical signal at a bit rate of 2.48832Gb/s by the E/0 18. In the
S/P & 0/H 28 this STS-48 signal is converted into byte (8-bit) parallel
form (311.04Mb/s on each of 8 lines) and is reframed, parity-checked,
and descrambled, and overhead information is extracted and supplied to
the control processor block 58. The S/P & 0/H 28 may further effect a
four-way sequential byte disinterleaving of the STS-48 signal, thereby
producing four parallel STS-12 signals at its output, each in byte
parallel form. This byte disinterleaving lowers the bit rate on
individual lines by a further factor of four, to 77.76Mb/s, to
facilitate signal processing, and is complemented by sequential byte
interleaving at appropriate points elsewhere in the node, for example
in the 0/H & P/S 38. However, for simplicity this is not further
discussed here.
The TSI 30 can assign each of the 48 STS-1 signals incoming to
it to any of 48 STS-1 signals outgoing from it, the assignments being
stored in a connection memory in the control processor block 58. The
same comment applies to each of the other TSIs 34, 42, 46, 52, and 54.
The synchronous nature of the STS signals facilitates this assignment
or time switching, so that this is readily achieved using conventional
25 time and/or space switching techniques. For example, conceptually for
each TSI the 48 incoming STS-1 signals could be separated onto 48
separate paths, and a 1-from-48 selector could be provided for each of
48 separate output paths to couple each output path to any input path,
the 48 output paths then being multiplexed as desired.
The SYNC 32 performs STS signal frame alignment by performing
STS pointer processing to align each of the 48 STS-1 signals to the
system frame. The synchronized STS signals are subject to
rearrangement in the TSI 34. The selector 36 selects STS-1 signals
from either this TSI 34 or the TSI 52, which interchanges STS-1 signals
derived from the node's receive port R, in dependence upon information
stored in memory in the control processor block 58, and supplies a
resulting STS-48 signal to the 0/H & P/S 3B. In the unit 38 overhead
~Z7't3~32
13
information from the control processor block 58 is multiplexed in the
SONET format, parity checking information is added, and the byte-
parallel STS-48 signal is scrambled and converted into a bit-serial
form for transmission via the E/O 20 and the outgoing port 14".
As indicated above, signals passing in the opposite direction,
from the port 16' and the receive port R towards the port 16", are
processed in a similar manner. In addition, STS-1 signals intended to
be dropped at the node's transmit port T are selected by the selector
56, also controlled by the control processor block 58, from the outputs
of the SYNCs 32 and 44.
In the node of Fig. 5, STS-1 signals incoming at the node's
receive port R are coupled via either the TSI 52 and selector 36 to the
outgoing port 14" of the path 14 or the TSI 54 and selector 48 to the
outgoing port 16" of the path 16. Signals incoming at the port 14' are
coupled via either the TSI 30 and the selector 56 to the node's
transmit port T or the TSIs 30 and 34 and the selector 36 to the
outgoing port 14". Signals incoming at the port 16' are similarly
coupled via either the TSI 42 and the selector 56 to the transmit port
T or the TSIs 42 and 46 and the selector 48 to the outgoing port 16".
For each STS-1 signal channel, the control processor block 58 maintains
in its connection memories appropriate information for routing the
signal channel via the node. Thus signal channels are routed through
the node via the TSIs and selectors to provide all of the desired
signal channel couplings between ports of the node as represented in
Figs. 3 and 4.
It will be noted from the above description that STS-1 signals
passing through the node between the ports 14' and 14" or between the
ports 16' and 16" in each case are conducted via two TSIs, 30 and 34 or
42 and 46. The description in this manner serves to facilitate
illustration of the node and an understanding of its operation, and the
provision of multiple TSIs and selectors as shown in Fig. 5 facilitates
the separation of the components of the node into modular units, such
as receiver units, transmitter units, and frame synchronization units,
which can be conveniently mounted in and interconnected by
communications equipment shelves in known manner. Alternatively,
however, the TSIs and selectors of the node may be rearranged to
9~32
14
provide the same functions in a different configuration possibly using
fewer TSIs and selectors.
In each node such as that shown in Fig. 5, the control processor
block 58 maintains in memory a connection map for each STS-1 signal.
In normal operation (Fig. 3) the information in this connection map is
used to set up the TSIs and selectors within the node to provide the
desired routing of each STS-1 signal. In a fault condition (Fig. 4)
protection routing information also in this connection map is used to
set up the alternative routing of each STS-1 signal affected by the
fault and terminated at the node. For fault detection, in each node
the control processor block 58 monitors each STS-1 signal individually,
and each STS-1 signal has its own overhead APS bytes K1 and K2. To
this end, the APS bytes are defined for protection purposes for each
individual STS-1 signal, although in the current SONET standard these
bytes are only used for STS-1 #1, i.e. the first STS-1 signal
multiplexed in a STS-n signal.
Referring again to Figs. 3 and 4, suppose that in the normal
condition of Fig. 3 a signal channel connection, i.e. an STS-1 signal
connection, exists from a head-end node, for example node 1, to a tail-
end node, for example node 3, via the path 14 and an intermediate node,for example node 2, with a corresponding connection in the opposite
direction via the path 16 from node 3 to node 1. In each direction and
hence on each path, this may for example be channel number 1 of the 48
channels available in the STS-48 signal. In the node 3, the control
processor block 58 monitors the line overhead information (e.g. 8-bit
interleaved parity information in SONET) for this STS-1 signal because
it is terminated at this node~ whereby it can detect transmission
errors for this signal. Accordingly, in the event of the interruption
between nodes 1 and 2 as shown in Fig. 4, the node 3 control processor
block is aware of this STS-1 signal being affected.
On detection of the fault, the control processor block 58 in
node 3 supplies, on the path 14 on a signal channel (e.g. the 48th STS-
1 of the STS-48 signal) which is available for a protection switch, the
head-end node number and the faulty channel number (in this case node
1, channel 1) in the APS overhead bytes K1 and K2 respectively. In the
node 17 these numbers are recognised by its control processor block 58
and a protection switch is made as shown in Fig. 4 to this protection
~;~79~3;Z
channel, number 48, on the path 16. When the node 3 control processor
block 58 recognises the channel 1 traffic in the protection channel
number 48, it completes the protection switch to the path 16 at the
tail-end. Similar protection switching takes place for the reverse
direction of transmission, and where necessary for each other STS-1
signal terminated at each node.
As described above, a node such as the node 3 which is not
adjacent a fault, such as the interruption between nodes 1 and 2, must
still be able to make a protection switch for connections terminated at
this node and affected by the fault. This requirement is avoided in an
alternative embodiment of the inventioll described below with reference
to Figs. 6 and 7.
The system of Fig. 6 is similar to that of Figs. 3 and 4, except
for the provision of loop-back paths within the nodes as described
below. Fig. 6 shows six nodes 1 to 6 coupled via oppositely directed
paths 14 and 16, and as in Fig. 4 shows an interruption in both paths
14 and 16 between the nodes 1 and 2.
As shown in Fig. 7, each node of the system of Fig. 6 comprises
the same elements as those described above with reference to Fig. 5.
In addition, each node includes two further TSIs 60 and 62, having
inputs coupled to the output of the S/P & 0/Hs 28 and 40 and outputs
coupled to third inputs of the selectors 48 and 36 respectively. The
TSIs 6~ and 62 are controlled by the control processor block 58 in a
similar manner to that already described above. Under this control,
the TSI 60 provides a loop-back path for STS-1 signals to be
selectively conducted from the incoming port 14' to the outgoing port
16", and the TSI 62 provides a loop-back path for STS-1 signals to be
selectively conducted from the incoming port 16' to the outgoing port
14". These loop-back paths are shown in Fig. 6 as additional solid
lines in the nodes 1 and 2 on each side of the fault, as compared to
the lines in these nodes in Fig. 4.
Fig. 6 also shows in dashed lines connection paths for a signal
channel connection between nodes 3 and 6 in the fault condition
represented. The line with longer dashes shows the path from node 3
port R to node 6 port T, and the line with shorter dashes shows the
reverse direction path from node 6 port R to node 3 port T.
~ ...... . . . .. ..
~'~79~32
16
It is assumed here that in a normal, non-fault, condition this
signal channel connection is from node 3 port R via path 16 and nodes 2
and 1 to node 6 port T, and from node 6 port R via path 14 and nodes 1
and 2 to node 3 port T. Thus this signal channel connection is
adversely affected by the fault between nodes 1 and 2. It is further
assumed here for purposes of explanation that in the normal condition
the signal channel connection is an STS-1 signal occupying channel
number 1 in each direction, i.e. channel 1 on the path 16 from node 3
to node 6 and channel 1 on the path 14 from node & to node 3, and that
channel 48 on each path, and hence in each direction of transmission,
is available for a protection switch. It is further assumed for
simplicity that there is no time slot interchanging of signals passing
through the nodes on the paths 14 and 16.
Now, in response to the fault between nodes 1 and 2, each of
these nodes on either side of the fault effects a loop-back connection
via the additional TSIs between the channel to be protected and the
protection channel. In other words, in the node 1 the control
processor block 58 controls the TSI 60 and selector 48 so that channel
1 on incoming port 14' is connected to channel 48 on outgoing port 16",
and channel 48 on incoming port 14' is connected to channel 1 on
outgoing port 16". Similarly, in the node 2 the control processor
block 58 controls the TSI 62 and selector 36 so that channel 1 on
incoming port 16' is connected to channel 48 on outgoing port 14", and
channel 48 on incoming port 16' is connected to channel 1 on outgoing
port 14". No switching is necessary for this signal channel connection
in any of the other nodes, whether these are the terminating nodes 3
and 6 or the intermediate nodes 4 and 5.
Thus in the fault condition shown, the connection from node 3
port R to node 6 port T is maintained via path 16 channel 1 to node 2,
via the TSI 62 in node 2, via path 14 channel 48 and nodes 3, 4, 5, and
6 in turn ~o node 1, via the TSI 60 in node 1, and via path 16 channel
1 to node 6. Conversely, the connection from node 6 port R to node 3
port T is maintained via path 14 channel 1 to node 1, via the TSI 60 in
node 1, via path 16 channel 48 and nodes 6, 5, 4, and 3 in turn to node
2, via the TSI &2 in node 2, and via path 14 channel 1 to node 3.
Thus in both the normal and the fault conditions, node 3
connects its port R to path 16 channel 1 and its port T to path 14
32
17
channel 1, and node 6 connects its port R to path 14 channel 1 and its
port T to path 16 channel 1. Hence neither node needs to take any
action in response to the fault, because all of the necessary
protection switching takes place in the nodes 1 and 2 on each side of
the fault. In this manner, the need for communication among the nodes,
other than nodes immediately adjacent the fault, for protection
switching purposes is eliminated. This advantage is achieved at the
cost of the addit;onal TSIs in each node and a greater path length over
which the signals are conducted in each direction in the fault
condition, but it does not increase the capacity required for
protection switching because for each transmission direction only the
original channel 1 and the protection channel 48 are used.
The above-described loop-back is effected in each node on either
side of a fault only for connections affected by the fault and not
terminated at the node. In the system of Fig. 6 a signal channel
connection between nodes 1 and 2 would be protected, without any loop-
back, in the same manner as in the system of Figs. 3 and 4, and a
signal channel connection between nodes 1 and 3 via node 2 would be
protected by loop-backs only in node 2.
The description above relates for simplicity and clarity to only
one double loop transmission system. Transmission systems, for example
telecommunications networks, are generally more complex than this and
desirably may have many adjacent and overlapping bidirectional loops.
The invention is equally applicable to the loops of such more complex
networks.
For example, Fig. 8 illustrates one node 70 of a communications
network, which node is at an intersection point of four looped
transmission systems each as described above with reference to Figs. 3
to 7. Thus in Fig. 8 transmission paths 141 and 161 correspond to the
paths 14 and 16 respectively of a first looped transmission system,
LOOP 1, as described above. Similarly, paths 142 and 162, 143 and 163,
and 144 and 164 correspond to the paths 14 and 16 respectively of other
looped transmission systems LOOP 2, LOOP 3, and LOOP 4. Although four
loops are referred to here, there may be an arbitrary number of
intersecting loops. The loops themselves may be adjacent one another,
overlapping, or within one another as desired for the particular
network.
. . . . . .: . ,
18
The network node 70 includes four node components 71 to 74, each
connected in a respective one of the looped transmission systems LOOP 1
to LOOP 4, each of which corresponds substantially to a node as
described above with reference to Fig. 5 or Fig. 7. The ports T and R
of each node component are coupled via paths, e.g. 81 and 91 for the
node component 71, to respective ports of a cross connect switch 76 of
the node 70. The cross connect switch 76 serves in known manner to
switch signals among its ports. Accordingly, signal channels in
different loops, for example LOOP 1 and LOOP 3, can be coupled via the
cross connect switch 76 to provide communications throughout the
network.
Obviously, within the network node 70 parts of the individual
node components 71 to 74 and of the cross connect switch 76 may be
merged with one another and in some cases (e.g. the interface circuit
26) simplified or dispensed with, whereby the network node as a whole
may be provided more conveniently and economically than the individual
components thereof illustrated separately for simplicity and clarity.
Although particular embodiments of the invention have been
described in detail, it should be appreciated that numerous
20 modifications, variations, and adaptations may be made thereto without
departing from the scope of the invention as defined in the claims.
