Note: Descriptions are shown in the official language in which they were submitted.
`~ 1 3338 1 8
SWITCHING NODE FOR A COMMUNICATIONS SWITCHING NETWORK
This invention relates to a switching node for a communications
network, and is particularly concerned with a switching node which can
be used to facilitate the communication of various types of information
having widely varying bandwidths and connection set-up and duration
requirements.
In a communications network, it is well known to provide
different types of switching nodes for different types of information.
For example, packet switches are used for switching relatively short-
duration data packets in a packet-switched communications network, and
circuit switches are used for switching relatively long-duration
information, such as voice or motion video communications, in a
circuit-switched communications network. Communications networks are
evolving as synchronous optical networks, in which information is
communicated via optical fiber communications paths between synchronous
- switching nodes. ATM (asynchronous time-division multiplex) switches
constitute a currently preferred form of packet or message switch,
which can operate conveniently in a synchronous optical network.
Message switches such as packet and ATM switches are particularly
well suited to handling reliably large numbers of relatively short
information messages, such as for example control information and
financial transactions, typically comprising up to about 103 bytes per
message. Thus the use of ATM switches is desirable for this type of
information in a synchronous optical network. However, such switches
are not well suited to handling much wider bandwidth and longer
duration messages, such as voice communications (telephone calls) which
may typically comprise 107 bytes per call or message, and motion video
communications which may comprise of the order of 101 bytes per
message. Conversely, circuit switches, which are well suited for
handling voice and video communications of relatively long duration,
are not well suited to handling large numbers of short messages due to
their relatively long connection set-up and release times, which may be
orders of magnitude greater than the duration of such short messages.
Especially with increasing numbers of LANs (local area networks)
and communications of data and software between them, and the provision
of other communications services such as facsimile and graphics
transmission, there is an increasing need for spontaneous
communications of information messages of the order of 104 to 106
~'
133381~3
bytes. For example, an operator at a data terminal may issue a command
to call up from a remote location a graphics screen comprising of the
order of 105 bytes; to respond to such a command without an undue delay
requires communication of the graphics screen information using a
relatively wide bandwidth connection which is rapidly set up. Circuit
switches do not readily satisfy the rapid connection set-up
requirement, and increasing the capacity of an ATM switch to handle
such messages presents problems relating to cell loss in ATM switches.
In particular, the bursty characteristics of such information in ATM
switch nodes may cause loss of short messages in the nodes even if
these are designed with a greatly increased capacity, and such losses
may cause a substantially complete disruption of communications via the
network.
Accordingly, an object of this invention is to provide an
improved switching node for a communications network for facilitating
the handling of these various types of message.
According to one aspect of this invention there is provided a
switching node comprising: circuit switching means for establishing
circuit-switched connection paths between inputs and outputs of the
switching node; message switching means for communicating messages via
the switching node; and control means arranged to control the circuit
switching means to establish the circuit-switched connection paths in
dependence upon connection signalling messages communicated via the
message switching means; wherein the control means comprises: means for
receiving a connection signalling message from the message switching
means; means for maintaining at least one queue identifying free
connection paths at the outputs of the switching node; means responsive
to an identification of a free connection path from said queue for
supplying an outgoing connection signalling message to the message
switching means in response to a received connection signalling
message; and means responsive to an identification of a free connection
path from said queue for establishing a connection via the circuit
switching means in response to a received connection signalling
message.
Thus in accordance with this invention short message traffic and
connection signalling messages for other, longer duration and higher
capacity traffic, can be handled by a message switch of relatively low
capacity and rapid response, while the longer duration and higher
_ 3 l 33381 8
capacity traffic is handled by a circuit switch controlled by the
signalling messages conducted via the message switch.
Conveniently the message switching means comprises an ATM
(asynchronous transfer mode, or asynchronous time division multiplex)
switch, which preferably is coupled to the circuit switching means for
conducting the connection signalling messages to and from the message
switching means via the circuit switching means.
According to another aspect this invention provides a switching
node for establishing message-switched and circuit-switched
communications in a switching network, the switching node comprising:
a circuit switch for establishing switched connections between inputs
and outputs of the switching node, the circuit switch including control
means; and an ATM message switch for establishing message-switched
communications via the switching node, the ATM message switch being
arranged for supplying to and receiving from the control means of the
circuit switch signalling messages for controlling the establishment of
said switched connections; wherein the circuit switch control means
comprises: means for receiving a signalling message from the message
switch; means responsive to a received signalling message for
establishing a connection via the circuit switch; means for maintaining
at least one queue identifying free connection paths at the outputs of
the switching node; and means responsive to an identification of a free
connection path from said queue for supplying to the message switch an
outgoing signalling message dependent upon the received signalling
message and the establishment of a connection via the circuit switch.
The invention will be further understood from the following
description with reference to the accompanying drawings, in which:
Fig. 1 is a chart representing characteristics of traffic in a
communications network;
Fig. 2 schematically illustrates a communications network
including a plurality of switching nodes;
Fig. 3 schematically illustrates a switching node in accordance
with an embodiment of this invention;
Fig. 4 illustrates in greater detail parts of one form of the
switching node of Fig. 3;
` 1 33381 8
Figs. 5 to 7 illustrate parts of alternative forms of the
switch;ng node; and
Fig. 8 illustrates parts of a preferred form of the switching
node.
Throughout the drawings, similar references are used in the
different figures to denote similar parts.
Referring to Fig. 1, there is illustrated a chart showing, for
various types of traffic, representative average busy-hour capacities
which could be required of a communications network carrying the
traffic. The horizontal scale is a logarithmic scale representing the
duration or length of a typical call in bytes for the respective type
of traffic, and the vertical scale is a logarithmic scale of total
traffic capacity in bytes/second for the respective type of traffic.
In order of increasing typical call length, i.e. from left to
right in Fig. 1, the columns of Fig. 1 represent the following traffic
types. In the following list and in Fig. 1, these traffic types are
also classified into three categories, namely short messages, file
transfers, and long calls:
Acknowledgements
Call Control Messages ) Short Messages
Financial Transactions
OA&M Messages
Software Transfer
Facsimile Transmission ) File Transfer
Graphics Transmission
Still Video Transmission
Voice Calls ) Long Calls
Video Transmission
It should be noted that the short messages and file transfers
are typically spontaneous types of traffic, in that it is desirable
that they be communicated without significant connection set-up
delays. In contrast, the long call traffic can tolerate connection
1 3338 1 8
set-up delays such as are present in existing voice call circuit
switches.
As can be seen from Fig. 1, short messages have typical call
lengths of less than 103 bytes, file transfers have typical call
lengths of 104 to 106 bytes, and long calls have typical call lengths
of 107 to 101 bytes. By far the greatest network capacity is
required for long calls, comprising telephone voice calls and video,
e.g. television, transmission, for which circuit-switched connections
are typically used in existing communications networks. On the
contrary, short messages, which are typically handled using existing
packet-switched communications, impose a relatively small requirement
on network capacity. As already discussed, file transfers require a
significant bandwidth or network capacity and also, like the short
messages, require short connection set-up times in view of their
spontaneous nature.
It has been considered as being desirable to handle all types
of traffic using a communications network with a single type of
switching node. In this respect, ATM (asynchronous transfer mode)
switches operating within a SONET (synchronous optical network)
optical communications network have been identified as a possible form
for such a network. However, the disparate traffic characteristics as
illustrated in Fig. 1 for the different categories of traffic indicate
that a different type of switching node may be preferable.
More particularly, the short message traffic in Fig. 1 is well
suited to being handled by ATM switching nodes, whereas the long call
traffic of Fig. 1 is better suited to being handled by STM
(synchronous time-division multiplex) switching nodes. Attempting to
handle long call traffic in ATM switching nodes leads to greatly
increased complexity for establishing semi-permanent virtual
connections through the nodes, and long call connection overhead and
set-up times in STM switching nodes make these undesirable for
handling the short message traffic. The intermediate, file transfer,
traffic can be handled by either type of switching node with
commensurate advantages and disadvantages.
Fig. 3 illustrates a different type of switching node in
accordance with this invention, which may be used in a communications
network as illustrated in Fig. 2.
_ 1 3~3~ ~ ~
Referring to Fig. 2, five switching nodes N1 to N5, each of
which can be of the form described below, are illustrated
interconnected by optical communications fibers 10. Only a few fibers
10 are shown in Fig. 2 for simplicity, and many more may be provided
to accommodate traffic requirements and opposite directions of
transmission. In particular, different numbers of fibers 10 may be
provided between different nodes according to traffic requirements;
for example two fibers 10 are shown between the nodes N2 and N3 for
carrying more traffic than can be accommodated on one fiber. Traffic
on each fiber 10 is time division multiplexed in conventional manner.
In the communications network of Fig. 2, traffic of arbitrary
type can be communicated between arbitrary nodes. For example, Fig. 2
represents by arrows 12 and 14 traffic incoming at the node N1 and
outgoing at the node N5; such traffic may for example be routed
through the network between the nodes N1 and N5 via the node N3 as
shown by an arrow 16. By way of example, this traffic is assumed here
- to comprise a signalling header followed by a data message of
arbitrary duration. As discussed in greater detail below, the
signalling header includes a network address of the destination for
the traffic, from which the switching nodes N1, N3, etc. determine an
appropriate routing for the traffic through the network, and may
include other information such as information concerning the nature of
the message (e.g. voice call, acknowledgement, etc.) and a check sum
for avoiding errors. Traffic of this general type is known for
example from an article entitled "Burst Switching - An Introduction"
by Stanford R. Amstutz, IEEE Communications Magazine, Nov. 1983, pp.
36 to 42.
Referring now to Fig. 3, there is illustrated a form of
switching node in accordance with an embodiment of this invention.
The switching node comprises an STM switch 20 and an ATM switch 22
which are associated with one another as described in detail below.
The STM switch 20 can comprise a conventional cross-point matrix or
time switch for establishing circuit-switched connections between
inputs 24 and outputs 26 under the control of a control part 28 of the
switch. The ATM switch 22 serves for establishing message-switched
connections (virtual circuits) or datagrams between its inputs and
outputs 30, and also can comprise a known form of switch. Examples of
1 333~ 1 ~
`_
the form of ATM switches are known for example from the following
articles:
"Asynchronous Time Division Techniques: an experimental
packet network integrating video communication" by A. Thomas, M.
Servel, and J.P. Coudreuse, ISS, May 1984;
"Integrated Wideband Networks Using Asynchronous Time
Division Techniques" by H. Le Bris and M. Servel, IEEE International
Conference on Communications, 1986; and
"Prelude: An Asynchronous Time-Division Switched Network" by
J.P. Coudreuse and M. Servel, IEEE International Conference on
Communications, 1987.
Inputs 24 of the STM switch 20 are coupled in known manner to
the fibers 10 (Fig. 2) incoming from other switching nodes, and
outputs 26 of the STM switch 20 are coupled in known manner to the
fibers 10 (Fig. 2) outgoing to other switching nodes. Links 32, 34,
36, and 38 are provided between the STM switch 20 and the ATM switch
22 for the communication of information therebetween, as described
further below.
Assuming for example that Fig. 3 represents the switching node
N3 of Fig. 2, then as illustrated in Fig. 3 an input 40 of the STM
switch 20 may be coupled to the node N1 and outputs 42 and 44 of the
STM switch 20 may be coupled to the nodes N5 and N1 respectively. The
traffic represented by the arrow 16 in Fig. 2 is routed through the
switching node N3 of Fig. 3, as shown by a long-dashed line through
the STM switch 20, in the manner described below.
As already mentioned above, the traffic incoming from the node
N1 on the line 40 comprises a signalling header and a subsequent
message (e.g. file transfer or long call traffic) of arbitrary
duration. More specifically, the signalling header is conveniently
provided on a pre-allocated signalling tdm channel on the line 40, in
the manner of common channel signalling, and the subsequent message is
provided on another tdm channel on the line 40. Thus on the tdm line
40 there can conveniently be one tdm channel pre-allocated for
carrying the signalling information for all of the other tdm channels
on this line. The STM switch 20 is set up (e.g. in known manner by a
management system of the network, not shown) to couple this signalling
channel information to the link 34 and thence to an input of the ATM
1 33381 8
-
switch 22, as shown by a short-dashed line in Fig. 3. Similarly,
signalling information on tdm channels at the other inputs 24 is
coupled to the ATM switch 22 via preset connection paths through the
STM switch 20 and via the link 34.
In a corresponding manner, signalling information produced by
the ATM switch 22 on the link 32 is coupled by preset connection paths
through the STM switch 20 to signalling tdm channels on the lines
connected to the outputs 26 of the STM switch 20, as shown in Fig. 3
for example by short-dashed lines through this switch 20 to the
outputs 42 and 44 for the nodes N5 and N1.
By means of these preset connection paths, the signalling
information for file transfer and long call traffic connections, and
likewise the short message traffic, can be routed to and from the ATM
switch 22 in each switching node via the same fibers 10 which are used
for coupling between the STM switches 20 in the switching nodes. This
also provides the advantage that the preset connection paths can be
modified by the network management system after initial set-up, to
suit particular requirements or operating conditions. However, it
should be appreciated that this arrangement is not essential, and that
the signalling information and short message traffic could
alternatively be coupled between ATM switches 22 in successive
switching nodes by paths which are distinct and separate from the
fibers 10 and which are not routed through the STM switches 20.
Accordingly, although as shown in Fig. 3 the links 32 and 34
extend between the switches 20 and 22, it should be appreciated that
they may instead be provided extending directly from the ATM switch 22
to the ATM switches in other switching nodes without passing through
the STM switch 20.
The signalling header and short message traffic which is
coupled to and switched by the ATM switch 22 is in the form of so-
called ATM cells, each of which has a standardized form which may
comprise a 5-byte ATM header and a 48-byte ATM payload or message.
The 5-byte ATM header contains an identity or address of an output of
the ATM switch 22 to which the message is to be coupled; the switch 22
monitors this address and routes the message accordingly. In the ATM
switch 22 of Fig. 3, the link 38 is allocated a respective ATM switch
output address so that ATM cells which contain signalling information
1 33381 8
-
for setting up connections via the associated STM switch 20 are
switched to it by the ATM switch 22. Such ATM cells are referred to
below as connection path requests, and as represented by the label REQ
in Fig. 3 are supplied to the control part 28 of the STM switch 20.
The control part 28 of the switch 20 also produces ATM cells
on the link 36, which are switched through the ATM switch 22 in a
similar manner to the link 32 to constitute either a positive
acknowledgement (ACK) constituted by a signalling information ATM cell
forwarded to the next node N5 or a negative acknowledgement (NACK)
constituted by an ATM cell returned to the previous node N1, as
further described below.
Fig. 3 illustrates by dashed lines in the ATM switch 22 the
routes taken through this switch for connection path requests from the
link 34 to the link 38 and positive and negative acknowledgements from
the link 36 to the link 32. In addition, Fig. 3 illustrates by a
dashed line 46 the coupling through the ATM switch 22 of short message
traffic; as this is handled, in ATM cells as discussed above, in known
manner it is not described further here. However, it should be
appreciated that this short message traffic, together with the
signalling information ATM cells described here, constitute a
relatively low capacity load which can be readily handled by the ATM
switch 22.
In the case of the signalling information ATM cells, each cell
contains in its ATM payload information which is necessary or
desirable for establishing and propagating a connection through the
STM switches or negatively acknowledging requests for connections
which can not be established. For example, in the case of a
signalling information ATM cell incoming (~.g. from the node N1) to
the ATM switch 20 (e.g. in the node N3) or forwarded as a positive
acknowledgement to a subsequent node (e.g. the next node N5), the ATM
payload of the cell can comprise such information as a call
identifier, the calling number, the identity of the node, port, and
channel originating the ATM cell, the destination network address,
optional bandwidth requirements and service type designations, and a
checksum. In the case of a negative acknowledgement, the ATM payload
can comprise the call identifier, calling number, identity of the node
producing the negative acknowledgement, originating port and channel
`- 1 33381 8
numbers, and a checksum. These various types of information are
given by way of example only, and not by way of limitation.
Thus the STM switch 20 is used to establish, under the control
of signalling information supplied via the ATM switch 22, circuit-
switched paths through the switching node for traffic having a high
capacity, for example the file transfer and long call categories in
Fig. 1. The control or signaling information for such circuit-
switched connections, and the short message category of traffic in
Fig. 1, is routed through the switching node using the ATM switch 22.
The particular division of traffic categories as discussed here is
given by way of example and may be varied for particular
circumstances. For example, some or all of the traffic in the file
transfer category could be switched using the ATM switch 22 alone, if
this has a suitable capacity.
In view of the alternative configuration of the links 32 and
34 as discussed above, in the following description with reference to
Fig. 4 the ATM switch 22 is shown as being connected only to the
control part 28 of the STM switch 20 via the links 36 and 38 for the
control of the circuit-switched connections.
Referring to Fig. 4, one form of the STM switch 20 and its
control part 28 is shown in more detail. The switch comprises a
matrix of cross-points 50, which are closed under the control of
connection memories 52 to form in known manner connections between
lines 54 incoming from previous nodes in the network and lines 56
outgoing to following or next nodes in the network. For convenience,
Fig. 4 illustrates one connection memory 52 associated with each
respective column of the cross-point matrix 50, and hence for each
line 56. Each memory 52 is a cyclic memory containing information as
to which cross-points are closed in each time slot of a tdm (time
division multiplex) frame.
In order to establish connections via the STM switch 20, the
control part 28 further includes a routing table 58, a connection path
request register 60, and a connection response register 62. The
routing table 58 contains stored information which identifies which
lines 56 are coupled to which other nodes of the network. Each
request for a connection through the STM switch 20 is supplied via the
ATM switch 22 and link 38 as described above to the request register
11 1 3338 1 8
60, and in accordance with the request and the information in the
routing table 58 a respective one of the connection memories 52 is
controlled to establish the desired connection through the cross-
points which it controls. This connection is established for a
particular output line 56 and in a particular time channel of the tdm
frame; the identities of these are supplied from the respective
connection memory 52 to the response register 62. As described in
greater detail below, the response register 62 assembles a new ATM
cell from this information supplied from the connection memory 58 and
the original cell information which is supplied from the request
register 60 via a line 61. The new ATM cell is supplied from the
response register 62 to the link 36 and is switched depending on its
ATM header, either as a forwarded signalling header to the next node
or as a negative acknowledgement to the previous node.
More particularly, as described above the incoming ATM cell
received in the request register 60 contains information identifying
the incoming port (line 54) and channel of the associated message to
be switched, and the destination network address. The destination
network address is used to determine from the routing table which
output line 56 is to be used for establishing the connection, thereby
identifying a respective connection memory 52 for establishing the
connection in known manner to the identified incoming port and
channel. Assuming that a time channel is free so that a connection
can be established, an identification of this time channel and the
output line 56 (and hence of the next node in the network for this
connection) are communicated to the response register 62. The
register 62 is thus able to produce, from this information and the
incoming ATM cell supplied via the line 61, an updated ATM cell with
an ATM header for routing to this next node and establishing the
connection therein. In this manner, the connection is established
progressively through the appropriate switching nodes until the
destination is reached.
In the event that, in any node, there is no free channel so
that the connection can not be established in this node, this is
communicated from the respective connection memory 52 to the response
register 62, which accordingly produces the negative acknowledgement
1 3338 1 8
12
ATM cell which is routed to the previous or originating node via the
ATM switch 22 to indicate the failure to establish the connection.
Whilst the switching node and network as so far described is
operable, it has certain practical disadvantages. In particular, the
registers 60 and 62 can accommodate only one connection request and
response at a time, and this involves a delay of up to one tdm frame
while the respective connection memory 52 determines whether or not
there is a free channel available, in addition to the time required
for processing the information. This delay would limit the number of
connection requests which could be handled to a maximum of one per tdm
frame, or 8000 per second in a typical network, which would not be
sufficient in practice. Furthermore, this delay would necessitate
some form of queuing of requests at the input to the request register
60. These disadvantages are substantially avoided by the modified
arrangements described below.
Fig. 5 illustrates an alternative and improved form of part of
the control part 28 of the STM switch 20. As shown in Fig. 5, a queue
70 of currently idle time slot channels for each respective one of
the lines 56 is maintained for example in a register or RAM (random
access memory), and hence for each connection memory 52, only one of
which is shown in Fig. 5 for simplicity.
In response to a connection path request causing the routing
table 58 to identify a particular line 56 and corresponding connection
memory 52 for establishing a connection via the cross-point matrix 50,
as indicated by a line 72 the identity of an idle channel is taken
from the top of the queue and is supplied immediately (via a line 74)
to the response register 62, and via a line 76 to the connection
memory 52. The immediate response to the register 62 from the queue
70 rather than from the connection memory 52 as in Fig. 4 enables the
register 62 immediately to produce the updated ATM cell, so that the
call signalling information is propagated rapidly through the
communications network, without incurr;ng the above delay in the
connection memory 52. Likewise, if no channel is free then the queue
70 is empty, so that this fact is immediately communicated to the
response register 62 to provide an immediate negative
acknowledgement.
13 l 33381 8
As illustrated in Fig. 5 by a line 78, when a channel becomes
free or idle at the end of a connection the connection memory 52
supplies its identity to the bottom of the queue 70. The detection of
a channel becoming free can be conveniently ach;eved in known manner
by flagging each busy channel and monitoring the presence of data on
the flagged channels, resetting the flag and the connection memory 52
and updating the queue 70 when data is no longer present.
In addition to providing a much more rapid response to the
register 62, whereby the above-described disadvantages are avoided,
this arrangement has the advantage that the channel which is
allocated for the connection is immediately removed from the queue, so
that many connection requests can be handled independently of the tdm
frame. Furthermore, the removal of free channels from the top of the
queue and their return to the bottom ensures that all of the channels
are used cyclically over time.
Consequently, it should be appreciated that this arrangement
greatly facilitates the handling of the file transfer traffic
described with reference to Fig. 1. By virtue of the rapid connection
path set-up as described above, a connection for such traffic can be
set up spontaneously without incurring conventional circuit-switched
connection set-up delays. In addition, this connection is set up as a
circuit-switched connection via the STM switch 20, which can readily
accommodate the desired bandwidth for the file transfer traffic, so
that the disadvantages of loading ATM switches with such traffic are
avoided.
Fig. 6 illustrates a further modification for the case, such
as between the nodes N2 and N3 is Fig. 2, where there are multiple
lines 56 leading to a single next node, any of which lines may be used
for a connection. As in Fig. 5, a queue 70 is maintained, but in Fig.
6 the queue identifies for each idle channel not only the identity of
the channel but also the identity of the associated connection memory
52 or line 56. The sharing of a single queue 70 among multiple lines
or facilities 56 provides an automatic selection and sharing or
distribution of traffic among these facilities.
Fig. 7 illustrates a further modification in which different
queues are provided for different types of traffic on the same output
line 56. For simplicity, Fig. 7 shows only the queues 70 and
`- 1 33381 ~
14
associated parts of the control part 28 for one line 56, and other
parts of the STM switch 20 are provided as in Figs. 5 and 6.
In the modified switch arrangement of Fig. 7, different time
slot channels on the line 56 may be allocated to different types of
traffic or services, for example leased circuit connections, network
administration, and central office trunks. A respective queue 70,
three of which are illustrated in Fig. 7, is maintained for each
different service type, channels being allocated from the top of each
queue for the respective service type under the control of the routing
table 58, and being returned to the bottom of the respective queue on
becoming free, individually in the same manner as described above.
To facilitate the allocation and queuing of channels, the connection
memory and/or queues may store in respect of each channel an
additional code identifying the service type to which the channel is
allocated. Such an arrangement also facilitates providing different
priorities for different types of traffic, in that higher priority
connection requests may be able to be fulfilled from more than one
queue.
The arrangement of Fig. 7 also facilitates providing a "busy
out" condition for individual channels and/or lines 56. For example,
if a subsequent node is inoperative, then all of the channels on the
line 56 leading to that node can be marked as busied out so that they
are never free. To this end, for example the three service types or
queues 70 in Fig. 7 can each be identified by a respective 2-bit code
which is used to address the queues 70. The fourth possible code
combination of such a 2-bit code can represent a busy-out state, which
is allocated to any channel which is not to be used, so that it is not
loaded into any of the queues 70.
Fig. 8 illustrates a preferred form in which the control part
28 of the STM switch 20 is implemented, and which may embody arbitrary
combinations of the arrangements of Figs. 4 to 7. To this end, the
control part 28 of the STM switch 20 comprises, in addition to the
cyclic connection memories 52, a RAM 80 and a micro-controller 82,
the latter communicating with the connection memories 52, the RAM 80,
and the ATM switch 22. In such a preferred arrangement the RAM 80
serves for storing information constituting the routing table 58, the
queues 70, and channel-service allocation information, and the micro-
1 3338 1 8
controller 82 operates in accordance with a set of programmedalgorithms to perform the functions of the registers 60 and 62 and for
communicating information between the RAM 80 and the connection
memories 52. In addition, in such an arrangement the information in
the RAM 80, and possibly also the algorithms in accordance with which
the micro-controller 82 operates, can be updated by communications via
the network itself from the network management system.
Although particular embodiments of the invention have been
described in detail, numerous modifications, variations, and
adaptations may be made thereto without departing from the scope of
the invention as defined in the claims.
