WO2010032214A2

WO2010032214A2 - Communication network

Info

Publication number: WO2010032214A2
Application number: PCT/IB2009/054096
Authority: WO
Inventors: David Denoon-Stevens
Original assignee: David Denoon-Stevens
Priority date: 2008-09-18
Filing date: 2009-09-18
Publication date: 2010-03-25
Also published as: EP2338243A4; WO2010032214A3; EP2338243A2

Abstract

A communications network is disclosed comprising a dynamic-impedance monitor- dynamic analysis network node 9 having multiple data ports (11, 18, 27), through which data is transmitted through the network. The impedances of external wire pairs 10.1, 10.2, 10.3 are monitored and respective termination impedances dynamically varied to minimize detected distortion. Signal flatness and reflected signal decay at the end of a data packet are used by a microprocessor as indicative of impedance matching or mismatching. The network further comprises "best route logic" for routing data through the node (9) to other data ports. If the network includes sufficient redundant links these are connected by the "best route logic" to form a second network over which status data can be transmitted.

Description

COMMUNICATION NETWORK

FIELD OF THE INVENTION THIS INVENTION relates to networks.

BACKGROUND TO THE INVENTION

In early computer systems, the topology comprised a central computer and a series of terminals connected to the central computer. The terminals had little processing power. Over time the topology of computer systems has changed. The central computer has been replaced by a series of individual computers connected together in a network.

In an office environment each individual computer can be a personal computer (desk top or lap top). In industry the personal computers of an office network are replaced by so-called intelligent information hosts (also called nodes) each of which has computing power and which are connected together in a network. The intelligent hosts "talk" to one another. They also receive information from, and send information to, local monitoring points. In data networks, the requisite network integrity is lost should a cable be damaged, resulting in an open circuit, should the network be shorted out (a

"hard" short) or there be a "soft" short problem where the cable changes its parameters. This can occur, for example, due to a partial short or moisture affecting the cable. In an office environment this is more of an inconvenience than a major problem. Some unsaved data may be lost but usually any problems which have arisen can be rectified without there being long term effects. However, when the network is that of a safety system such as a fire detection system or a manufacturing plant control network then the results can be more serious. An auto selecting redundant alternative communication path is then desirable.

Networks which are "self healing" have been proposed. Simply by way of example a system is known in which a communications node has two or more paths to connect it to its adjacent neighbour. If signalling is compromised on one path then the node uses another path to connect to its neighbour.

Once a node receives data through one of its ports, it passes it on to its other port, if it has two ports, or to one or more of its other ports if it has more than two ports. In this manner data is spread throughout a network.

The adjacent nodes are structured to detect the quality of the data links between them and, in the event of a fault which prevents data being transmitted along a specific route, an alternative route is chosen by the two connected nodes.

An object of the present invention is to provide a network in which redundant communications paths are utilized more effectively then in known networks. Another problem with current networks in that communications cable has impedance and the terminations at each end of the cable must have matching impedances to prevent reflected signals from distorting the data being transmitted. At installation impedances are used to terminate the cable but the impedances used are not necessarily correct. Changes in cable parameters over time can result in the impedances used becoming even less suitable.

The present invention seeks to improve network operation through correct cable termination even if non ideal cable is used as the communication cable and the redundant link paths.

BRIEF DESCRIPTION OF THE INVENTION

According to one aspect of the present invention there is provided a network comprising first and second nodes connected by an external wire link, each node including means for monitoring signals transmitted to it from the other node and for detecting distortions of the received signal and means for adjusting the termination impedance in dependence on the detected distortion of the received signal.

The distortion detecting means can detect the deviation from flatness of the signal and monitor signal decay time at the end of a data packet. Each node can include means for storing information pertaining to expected data format parameters, and means for isolating incoming data that does not conform to such format.

It is also possible to store information pertaining to signal strength and data quality.

Each node can have two or more ports for connection to external wire links, there being non-galvanic isolators such as optocouplers and inductive couplers through which signals received at the ports are transmitted to other ports.

According to another aspect of the present invention there is provided a network comprising a plurality of nodes which are connected to one another by data links forming a primary network and redundant links forming a secondary network, data being transmitted over the primary network and status information over the secondary network, each node including logic software for transferring data transmission from a data link of the primary network to a redundant link of the secondary network in the event that the integrity of signal transmissions over the data link falls below a predetermined threshold.

A table can be stored electrically in the node, the table containing the network designer's directives in respect of preferred active links. Said logic software can be designed to overrule the designer's directives in the event of a communications failure.

The network can include means for permitting transmission of data to be periodically transferred from all or part of the primary network to all or part of the secondary network and means for testing the cabling of the primary network by transmitting test signals over such cabling at an increasing Baud rate until a deterioration of data signal quality is detected.

The network can further include means for generating a report signal indicative of detected signal transmission performance of the primary network.

To reduce the effects equipment failure on the integrity of the network, each node can include a local port for connection to a data source, each local port being connected to a port of another node whereby, upon failure of the first mentioned node, data is transmitted to the port of said another node.

According to a further aspect of the present invention there is provided a network node including means for detecting signal strength and signal quality, and a series of light sources which are illuminated in dependance on the detected signal strength and signal quality to provide a visual indication of network status. The lights can be of various colours.

According to another aspect of the present invention there is provided a method of setting up a network which comprises connecting two nodes to one another by means of an external wire link, transmitting signals of known characteristics between the nodes, detecting signal distortion and adjusting the termination impedance to reduce signal distortion and the end of data packet signal decay time.

The method can comprise transmitting long and short signals, the long signal acting as a reference and the difference between the detected heights of the long signals and the short signals being used to determine necessary impedance changes.

BRIEF DESCRIPTION OF THE DRAWINGS

For a better understanding of the present invention, and to show how the same may be carried into effect, reference will now be made, by way of example, to the accompanying drawings in which:-

Figure 1 is a diagram of a dynamic impedance monitor - dynamic network analyzer network node;

Figure 2 illustrates a simple non-redundant dynamic impedance monitor external link including two of the nodes of Figure 1 designated DIM-DNA;

Figure 3 illustrates a redundant chain network topology; Figure 4 illustrates a redundant star network topology; Figure 5 illustrates a redundant star - chain network topology which includes a remote monitoring connection link;

Figure 6 illustrates a redundant trunk wire network topology; and Figure 7, 8 and 9 illustrate wave forms.

DETAILED DESCRIPTION OF THE DRAWINGS

Figure 1 illustrates a dynamic impedance monitor - dynamic network analysis network node 9. The node 9 has multiple data ports 11 , 18 and 27, the data coming in on one data port being communicated to, and then retransmitted by, other data ports. By this means data is distributed throughout the network of which the node 9 forms a part. The node 9 also terminates the external communications wires constituting the communications links 10.1 , 10.2, 10.3 in an optimal manner as will be described. An information source is connected to the network through a local port 33 of the node 9. There can be more than one information source connected to the node. The data from the local information source or sources is re-transmitted by other ports of the node 9. Local information sources providing data could, for example, be fire or smote detectors when the network is a fire detection system or temperature, flow rate or pressure detectors in an industrial plant. The local information source can also be a program transmitting instructions to control an industrial process.

The data source connected to the local port 33 can also be connected to one of the ports of another node 9. If this topography is used, failure of one of the nodes 9 does not result in the data from the source connected to port 33 being lost. The data will be accepted by the other note to which the data source is connected and be distributed through the network. This ensures that the system is equipment fault tolerant as well as cable fault tolerant.

The node 9 illustrated in Figure 1 is connected to external wire pairs constituting the communications links 10.1 , 10.2, 10.3, by means of the node ports 11 ,18 and 27.

The node 9 monitors the external wire pairs by means of the link ends designated 12, 20 and 26 and selects a correct termination impedance to terminate the wire pairs. This allows for the use of non ideal wire for some or all external data links without encountering the problems usually associated with non-ideal communication cable.

The methodology to be described allows a matched termination impedance to be selected which yields a low reflected signal. Such signals are sometimes called "echoes". If there is a mismatch between the impedance of the external wire pairs constituting the links 10.1 , 10.2 etc and the impedance of the link ends 12, 20, 26, reflections or echoes manifest themselves and these reflections create false data signals which can interfere with data signalling and increase the bit error rate. The bit error rate is usually shortened to BER. The effects of cable reflection noise on the line are eliminated by impedance matching, or at least minimized, when the termination impedance matches the external communications link's "characteristic" impedance.

When the network node's external wire links are set up, the dynamic impedance monitors 12, 20, 26 terminate the link at a nominal value selected by means of a digitally selectable impedance device. A standard known data pulse of long and short data transitions is sent down the external link wire. The monitor takes many analogue to digital samples while the pulse train is being sent thereby to monitor for signal distortion or signal reflections during and at the end of transmission. In Figure 7 the square originally sent signal is received as a square received signal. This indicates impedance matching. In Figure 8 the signal is received in distorted form due to mismatched cable link and termination impedances. If there is a termination mismatch then the signal does not transition from low to high as a square level change but the signal has a slope on the high or low section (Figure 9).

The monitors 12, 12, 26 change the termination impedance and then resend the known data pulse train until the best flattest signalling pattern with the least signal reflection is observed. Once the monitors at each end of the data link are setup then the monitors set the link into run mode and they store these values for run time referencing purposes. During run time when similar signalling patterns are observed then wave shape comparisons are made with the stored shape patterns and the deviation of signal quality can be detected and reported. When a link is carrying data no further status information is sent over the data channel as the data channel is dedicated exclusively to the network information sources data. Further status information can be sent if there are redundant cable links that can be used for dispersing network status information as will be described below. A fault condition is indicated on the status indicators 16 and is used to activate a relay 24 which in turn transmits the fault warning to be a remote location.

To perform these impedance adjustments and data analysis, a small microprocessor is typically provided at each matched termination block 12,20,26

These microprocessors, while the data channel is running, monitor the signal flatness and the reflected signal decay at the end of data packet and make small impedance adjustments. If the signal shape is improved by these small incremental adjustments then they are retained.

Instead of using a plurality of microprocessors it is possible to use a single central processor. In this form analogue readings are taken by dedicated analogue to digital converters and the digital information is sent to the central processor. The central processor makes the necessary impedance changes.

If there is a status data channel available at the network node then the impedance adjustment can be logged as well as the deviation from start up. From this information it can be determined if there is a continuous variation in Impedance in one direction. This can be used to trigger an alert situation and ensure remedial action before the link deteriorates into a non-working data link. If there are redundant links of acceptable characteristics links between nodes then periodically the active data channel can be switched over to one of these links. The primary data link can then be checked for signalling quality with the known data pulse train of test signals. Data signalling quality can be determined and the link can be corrected with a new matching impendance value. One specific test method comprises increasing the Baud rate until a deterioration of the data signal quality is detected.

In embodiments of this invention which use differential signalling, eg with USB2 signalling where one signal conductor goes positive and other associated signal conductor goes negative, the termination Impedance at the end of the copper cable may be constituted by three separate digitally selectable termination resistors which together are referred as the termination impedance.

The primary termination resistor is connected between the two conductors. One of two higher termination impedance values is connected from one conductor to the positive supply voltage. The other high impedance is connected from the other conductor to the zero volts supply. Each of these Impedances is changed and the flatness criteria and signal decay at the end of transmission is monitored. If the quality of the signal improves the adjusted value is retained. Monitoring and maintaining the correct impedance is performed on a continuous basis. Through this methodology any daily or seasonal changes in cable parameters are tracked and the correct matching termination impedance is maintained.

The external ports 11 ,18, 27, 33 are isolated electrically from each other by means of non-galvanic isolators 14,22,30,24 such as optocouplers or inductive couplers. Data can consequently be shared across the isolators but electrical isolation is maintained between links 10.1 , 10.2, 10.3 and 19.

It is well known that adding cable to existing systems causes problems due to the additive loading effects of the additional cable. This problem is avoided if the cable segments are isolated and terminated as described herein.

The block 15 is herein referred to as a dynamic network analyzer data router block and includes "best route logic" responsible for routing incoming data to other active data ports.

The format of the incoming data is checked against the expected data baud rate and data byte format settings set on the switch setting switches 17. These switch settings can be physical switches or stored software data parameter settings. If the data coming in at one port is detected by the serial RX/TX blocks 13, 21 , 25, 31 to be at a wrong baud rate or the data format is incorrect then the dynamic network analyzer block 15 isolates this non-conforming data. In this way data integrity across the network is maintained as information sources that are not conforming to the desired system communication parameters are isolated. Similarly the data signalling information is passed into the dynamic network analyzer data router block. This together with the digital analysis information described above is factored into "the best route" logic data link analysis and the link data quality reporting. A data format fault is signalled on status lights 16 and transmitted on the status network unless transfer of links from the status network to the data network has undermined the integrity of the status network.

If any port 11 ,18, 27, 22 is used as a connection to an offsite reporting terminal then the network status and data signalling quality in each external link can be ascertained from a remote location. This allows remote network diagnostics to be performed, provides the ability to perform "tele" servicing and permits the creation of system service reports at the remote location.

The external "wired" links 10.1 ,10.2 ,10.3 may not be entirely of one material such as copper but maybe include other communications media. For example, there can be a copper to optic fibre converter or a copper to radio link converter. These are typically converted back to copper wire before the next communications network node is reached. The nodes 9 correctly terminate the copper wires at both ends of the external link.

The node 9 is supplied with power along two separate power supply paths which lead to a dual power supply monitor 35. Redundancy in the supply of power is consequently also provided for in the illustrated communication system in that failure of one power supply path does not compromise the ability of the system to function.

The monitor 35 is itself monitored by the block 15 and the status of the power supplies is displayed on the monitor 35.

If there is cabling available which can be used to form a secondary network, the status of the power supplies can be transmitted to a remote location.

The signals generated by the node 9 and pertaining to signal qualities and data transmission quality can be used to produce a certificate of signalling performance which can be stored, and/or transmitted to remote locations and printed for submission to the regulatory authorities in compliance with statutory and other requirements.

It is possible with the communication network disclosed for cabling to be tested to ensure that it is in operational condition. This can be done by increasing the

Bauch rate incrementally on a redundant link of the secondary network until a deterioration in data signal quality is observed. If the results are stored it is possible, in subsequent tests, to determine if the properties of the cable have changed since the previous test.

It will be understood that each cable has specified nominal impedance and it is this impedance that is used during set up to match the cable and the termination impedances. The values of the impedances used can be stored so that it is possible to verify that the impedances used are of the correct value.

Acceptable upper and lower impedance values can be set for each cable. Should an adjusted terminal impedance be detected as being out of the applicable range, an alarm condition can be established warning of that cable's deterioration.

In Figure 2, a non-redundant external link 10.1 comprising a pair of preferably twisted wires is shown connecting two nodes 9.1 , 9.2. The nodes 9.1 , 9.2 in the illustrated form have information sources 8.1 and 8.2 which are connected to one another by means of the nodes 9 and the link 10.1. A problem with the link 10.1 would be indicated by the status lights 16 of Figure 1.

In Figure 3 the network comprises a redundant chain topology. There are at least two external wire links connecting the nodes 9 together. There is an extra link linking nodes 9.1 and 9.2 together through nodes 9.3 and 9.4. This caters for a power loss redundancy meaning that if any one unit loses power, data signalling is not hindered. Each node 9.1 , 9.2 has an information source 8.1 , 8.2, etc. connected to it.

The redundant spare external links between the nodes are used to provide a second status network. On this status network status information from each node 9 is transmitted to the other nodes. This status information can be passed on to a remote link for remote diagnostic viewing or for "tele" servicing purposes.

In Figure 4 the network comprises a redundant star topology. Each node 9.1 , 9.2, etc. has a direct link with each other node as well as an indirect link.

In Figure 5 the network comprises a redundant star - chain topology. Some clusters of nodes have direct star links and other nodes have redundant chain connections. A connection from node 9.6 to an external remote monitor 50 is shown at 52.

In Figure 6 the network comprises a redundant trunk topology. In this form there are two pairs of common data cables 8.1 , 8.2 connecting the nodes together. The redundant cabling is simple but the segment isolation feature of other topologies such as that of Figure 3 is lost through this simplified cable topology.

If there are at least three external link cables as shown in the Figure 3, Figure 4 and Figure 5 wiring topologies , then one cable can be damaged and data and network status continues on the remaining external links.

The preferences of an industrial network designer are stored as a table in the block 15 as directives. These directives may be a cable route due to physical security considerations e.g. primary cable layout preferred route or a cable route preference based on power supply. These directives will be followed unless a data link is measured as, say, 25% worse than a redundant backup secondary link. The "best route logic" is empowered to make the best data channel selection to keep the network up if a data cable becomes defective. The "best route logic" takes the network designer's directives into account but it has the power to override these if an alternative data link is available and is not already in use as a data link. Similarly the second data route preference directive from the network designer is also followed by the "best route logic". Where there are network designer preferences for a port not to terminate, as is the case in Figure 6 in respect of nodes 9.2 and 9.3 which are not at the ends of the cable, the "best route logic will adhere to this directive. However, if the signal changes due to cable breakage the network node would detect the signal distortion and override its directive in seeking to bring correct termination.

If the system designer does not store system design preferences in the "best route logic" the " best route logic" will allocate primary/secondary data and status channels. Where a node 9 has multiple external links a microprocessor powers the "best route software". The "best route logic" logically connects all redundant data links and this creates a second status network to transmit status data throughout the network if there are enough redundant links available.

Claims

1. A network comprising first and second nodes connected by an external wire link, each node including means for monitoring signals transmitted to it from the other node and for detecting distortions of the received signal and means for adjusting the termination impedance in dependence on the detected distortion of the received signal.

2. A network as claimed in claim 1 , wherein the distortion detecting means detects the deviation from flatness of the signal and monitors signal decay time at the end of a data packet.

3. A network as claimed in claim 1 or 2, wherein each node includes means for storing information pertaining to expected data parameters, and means for isolating incoming data that does not conform to such format.

4. A network as claimed in any preceding claim, wherein each node has two or more ports for connection to external wire links, there being non-galvanic isolators such as optocouplers and inductive couplers through which signals received at the ports are transmitted to other ports.

5. A network comprising a plurality of nodes which are connected to one another by data links forming a primary network and redundant links forming a secondary network, data being transmitted over the primary network and status information over the secondary network, each node including logic software for transferring data transmission from a data link of the primary network to a redundant link of the secondary network in the event that the integrity of signal transmissions over the data link falls below a predetermined threshold.

6. A network as claimed in claim 5, wherein a table is stored electrically in the node, the table containing the network designer's directives in respect of preferred active links.

7. A network as claimed in claim 6, wherein said logic software is designed to overrule the designer's directives in the event of a communications failure.

8. A network as claimed in any one of claims 5 to 7 and including means for permitting transmission of data to be periodically transferred from all or part of the primary network to all or part of the secondary network and means for testing the cabling of the primary network by transmitting test signals over such cabling at an increasing Baud rate until a deterioration of data signal quality is detected.

9. A network as claimed in claim 5, 6, 7 or 8 and including means for generating a report signal indicative of detected signal transmission performance of the primary network.

10. A network as claimed in any one of claims 5 to 9, wherein each node includes a local port for connection to a data source, each local port being connected to a port of another node whereby, upon failure of the first mentioned node, data is transmitted to the port of said another node.

11. A network node including means for detecting signal strength and signal quality, and a series of light sources which are illuminated in dependance on the detected signal strength and signal quality to provide a visual indication of network status.

12. A network node as claimed in claim 11 , wherein lights are of various colours.

13. A method of setting up a network which comprises connecting two nodes to one another by means of an external wire link, transmitting signals of known characteristics between the nodes, detecting signal distortion and adjusting the termination impedance to reduce signal distortion.

14. A method as claimed in claim 13 which comprises transmitting long and short signals, the long signal acting as a reference and the difference between the detected heights of the long signals and the short signals being used to determine necessary impedance changes.