MXPA98003956A - System of management of failure for a telecommunication network - Google Patents

Abstract

The present invention relates to a method of operation in a fault management system for an access network that is part of a public telecommunication network. In the access network, the terminal lines in the form of pairs of copper wires extend from a local switch (10) through a series of nodes to the terminal equipment provided for the network user. The fault management system includes a test hydrostatic head (104) and an access network management system (102). Each night, the hydrostatic test height (104) executes a series of tests on each of the terminal lines. The results of the test are transmitted to the access network management (102). The test results are then converted into circuit registers, each of which is indicative of the operational quality of the tested circuit. For each node, the circuit logs of the tested circuits passing through the node are combined to produce a node register that is indicative of the operational quality of the node. In order to identify the node that has the worst operational quality and therefore that has the greatest need for research, the nodes are classified according to their records of no

Description

SYSTEM OF MANAGEMENT OF FAILURE FOR A TELECOMMUNICATIONS NETWORK DESCRIPTION OF THE INVENTION This invention relates to a fault management system for handling faults in the terminal circuits of a telecommunications network and also to a method of operation of such a fault management system. A conventional public telecommunications network comprises a relatively small number of interconnected main switches and a much larger number of local switches, each of which is connected to one or two of the main switches. The local switches are connected to the terminal circuits of the network and the far ends of those circuits are connected to the terminal equipment as • telephone instruments provided for the users of the network. The network formed from the main switches and the local switches is known as a core network, while a network formed from the terminal circuits is known as an access network or local circuit. In this specification, it will be referred to as an access network. Some terminal circuits are connected to a remote concentrator, which may or may not have communication capability. The remote concentrator is then "connected to a local switch." In this specification, the thermal "local switch" is to be interpreted to cover local switches and remote concentrators.In a conventional access network, each terminal circuit is formed of a pair of copper wires Typically, each pair of copper wires passes through a series of nodes between the local switch and the terminal equipment Examples of such nodes are primary cross-connection points, secondary cross-connection points, distribution and junctions Recently, optical fibers have been used to transport terminal circuits in access networks.In a modern access network, both pairs of copper wires and optical fibers are used to transport the terminal circuits.When a terminal circuit is transported by an optical fiber, the circuit will typically pass through several nodes between the local switch and the computer. In each node, the fiber that enters from the local switch is deviated within a group of fibers that branch in several directions. When a terminal circuit is transported by an optical fiber from the local switch, the last part of the circuit can be transported by a pair of copper wires. Unfortunately, the terminal circuits are prone to faults. In a conventional access network, examples of such failures are disconnection, a short circuit between the two wires of a pair of wires, and a short circuit between one of the wires and ground. The causes of the faults include the entry of water into the node and also the physical damage to a node. The local switches are provided with circuit testers that can be used to test their terminal circuits. When a customer reports a fault in a terminal circuit, the circuit can be "• * tested later to identify the fault condition. To repair the fault, it is the current practice for an engineer to make an assumption on the location of the fault and open the node at that location. If the engineer's assumption is incorrect, you will have to open one or more additional nodes before finding and fixing the fault. Sometimes an engineer will have to open each node in turn between the local switch and the terminal equipment until it finds a node where the fault occurs. Typically, in a conventional access network, an engineer will open between 2.5 and 3 nodes on average before locating a fault. The present practice to locate faults therefore suffers from two problems. First, it takes time for an engineer to have to open several nodes before locating and fixing a fault. Second, since the nodes are of a delicate construction, every time an engineer opens a node, it will damage the node as a result of another failure in a terminal circuit. Fault management is described in an article entitled "An Integrated ISDN fault management system", Schimazake et al, Globecom '90, Session 802, page 7, volume 3, December 2, 1990, San Diego USA, pages-1503 to 1507 This fault management system is a waiting system that collects fault messages from network elements and analyzes them to find fault locations.

In accordance with one aspect of this invention, there is provided a fault management system for a telecommunications network that includes a local switch and a set of terminal circuits that extend between the local switch and the equipment provided to the users of the network. network, each of the terminal circuits passing through a series of nodes between the local switch and its respective terminal equipment, the fault management system comprising: circuit test apparatus located in the local switch and placed to execute circuit tests in the terminal circuits; a storage containing data related to the terminal circuits and such nodes; means for instructing the circuit test apparatus to execute a test suite on one of the terminal circuits; means for verifying the results of a set of tests executed by the circuit test apparatus by the presence of an expected fault, the means of checking that are placed to produce a fault report when a failure is assumed; means for identifying the nodes of a terminal circuit in which a present fault is suspected; means to evaluate a record of each node in a terminal circuit line in which a failure is suspected that represents the probability that the alleged failure is present in the node, the means of evaluation using the failure report in relation to the suspected failure and the data contained in the data storage to evaluate the record; and means for classifying nodes of a terminal circuit on which a failure is suspected in accordance with its records found by the means of evaluation; therefore, in use, after testing a circuit in which a failure is suspected, a list of the nodes is produced in such a circuit in which the nodes are classified according to the proof that the fault is present. in each node. With this invention, as a result of the classification of the nodes in a terminal circuit over which a failure is suspected according to the probability that the fault is present in each node, an engineer can be directed towards the node in which the fault is more likely to be present. With this invention the engineer is not directed to the correct node on each occasion, the present invention reduces the number of nodes that an engineer has to open before locating the faults. On average, when the invention is used in a conventional access network, it is estimated that an engineer will need to open less than 1.5 nodes before finding the node in which the fault is present. According to the second aspect of the invention, there is provided a fault management system for a telecommunications network system, in the telecommunications network that includes a local switch and a set of terminal circuits that extend between the local switch and the terminal equipment provided for the users of the network, each of the terminal circuits passes through a series of nodes between the local switch and its respective terminal equipment, the fault management system comprises: the circuit test apparatus located in the local switch and placed to execute circuit tests in the terminal circuits; a computer system for operating the circuit test apparatus; the computer system includes a storage containing data related to the terminal circuits and the nodes; The computer system is controlled by at least one program to execute the following operations: indicate to the circuit test apparatus that it executes a set of tests on one of the terminal circuits; verify the results of a set of tests performed by the circuit test apparatus for the presence of an assumed fault and produce a fault report when a supposed fault is present; identify the nodes of a terminal circuit over which a supposed fault is present; - evaluate a »-circuit of each node in a terminal circuit over -which is suspected of having a fault representing the probability that the assumed fault is present in the node using the fault report related to the assumed failure and the data contained in storage; and classify the nodes of a terminal circuit in which a failure is suspected according to their records. According to a third aspect of this invention, there is provided a method of operation of a fault management system for a telecommunications network, the telecommunications network includes a local switch and a set of terminal circuits that extend between the local switch and the terminal equipment provided for the users of the network, each of the terminal circuits passing through a series of nodes between the local switch and its respective terminal equipment, the fault management system comprising: circuit test apparatus located in the local switch and placed to execute circuit tests on the terminal circuits; and a computer system for controlling the circuit test apparatus, the computer system includes a storage containing data related to the terminal circuits and the nodes; the method comprises the following steps executed by the computer system: instructing the circuit test apparatus to execute a series of tests on one of the terminal circuits; verifying the results of the set of tests executed by the circuit test apparatus for the presence of an assumed fault and producing a fault report when an assumed fault is present; when a failure is suspected, identify the nodes of the terminal circuit on which it is suspected that it is failing; evaluate a record for each of the nodes of the terminal circuit in which a failure is suspected, which represents the probability that the assumed failure is present in the node based on the failure report related to the assumed failure and use the data contained in the storage of data; and classify the nodes in the terminal line on which the assumed failure according to the records found in the record evaluation stage.

This invention will now be described in greater detail, by way of example, with reference to the drawings in which: Figure 1 is a block diagram of an associated access network and local switch that is part of a telecommunications network in which the present invention can be used; Figure 2 is a block diagram showing the arrangement of some of the local switches and main switches of the telecommunications network mentioned with reference to Figure 1; Figure 3 is a block diagram showing the components of the telecommunications network that are used to provide a fault management system that modalizes this invention for the access network of Figure 1; Figure 4 is a block diagram of the main electromechanical components of a typical computer; »-. 'Figure 5 is a flow chart of the stages that are executed in the management system of a Figure 6 is a flow diagram of the stages that are executed in the fault management system in the estimation of the location of a supposed fault in the access network; Figure 7 is a functional block diagram of the fault management system; Figure 8 is a circuit diagram illustrating some of the measurements that are made when testing a terminal circuit; Figure 9 is a flow chart of the steps that are executed in the fault management system to monitor the operational condition of the access network; Figure 10 is a graph illustrating how the resistance value is converted into a converted value when one of the stages shown in the flow chart of Figure 9 is executed, and Figure 11 is a graph illustrating some of the results Experiments obtained by monitoring "the nodes using the stages shown in the flow chart" of Figure 9. Referring now to Figure 1, a local switch 10 and a conventional access network 12 connected to the local switch 10 are shown. The local switch 10 and the access network 12 form part of a public telecommunications network. The local switch 10 is connected to the terminal circuits or lines of the access network 12. Typically, a local switch is connected to several thousand terminal circuits. Each terminal circuit or line passes through several nodes before reaching its respective terminal equipment. These nodes comprise primary cross-connection points, secondary cross-connection points, distribution points and junctions and examples of those nodes will be described below. In the conventional access network 12 shown in Figure 1, each terminal circuit or line is formed from a pair of copper wires. The copper wires leave the local commutator 10 in the shape of one or more cables. One of these cables is shown in Figure 1 and indicated by the reference number 14. The remote end of the cable 14 from the switch 10 is connected to a primary cross-connection point 16 which may be housed in a street cabinet or underground junction box. From the primary cross connection point 16, the terminal lines branch like cables in various directions. For simplicity / in Figure 1 only three cables 18, 20 and 22 are shown. The remote end of the cable 18 is connected to a junction 19. The junction 19 is connected by the cable 21 to a secondary cross connection point 24. Distant ends of cables 20 and 22 are connected, respectively, to points <the secondary cross-connect connection 26 and 28. For reasons of simplicity, the continuations of the termination lines beyond the secondary cross connection points 24 and 26 are not shown. The secondary cross connection points 24, 26 and 28 are housed in junction boxes that may be located above or below the ground. From the secondary cross-connection point 28, the terminal lines are branched again in several directions in the form of cables. By way of illustration, Figure 1 shows the cables 40, 42 and 44 emerging from the secondary cross connection point 28. The cables 40 and 44 are connected respectively to junctions 46 and 48. The connections 46 and * 48 are connected , respectively, to the cables 50 and 52, the remote ends of which are connected to the distribution points 54 and 56. The remote end of the cable 42 is connected to a junction 60, the junction 60 is connected by a cable 62 to a distribution point 64. For reasons of simplicity, the terminal lines beyond distribution points 54 and 56 are not shown. The distribution points are implemented as junction boxes that are typically located on telephone poles. From each distribution point, the terminal lines branch out as individual copper wire pairs to where the terminal equipment provided for a network user is located. By way of illustration, Figure 1 shows two pairs of individual copper wire 70, 72 leaving the distribution point 64. The far ends of the copper wire pairs 70 and 72 are respectively connected to the terminal equipment 74, 76. As is well known, terminal equipment can take several forms, for example the terminal equipment can be a public telephone located in a public telephone booth ,, ... a telephone instrument located in a particular domicile or an office, or a fax machine or a computer located at the customer's premises. In the example shown in Figure 1, each of the junctions 19, 46, 48 and 60 is used to connect two cables together. The joints can also be used to connect two or more smaller cables to a larger cable. The cable 14 is housed in a duct. The air in the cable 14 is maintained at a pressure above ambient pressure. These • inhibit the ingress of water into the cable 14. On each terminal line, the two wires of each pair are designated as wire A and wire B. On the local switch 10, to supply the current to the line, a voltage deviation 50V is applied between the wire A and the wire B. As the deviation voltage is applied in the first changes using a battery, the deviation voltage is still known as the battery voltage. In the terminal equipment, wire A and wire B are connected by a capacitor, the presence of which can be detected when the terminal equipment is not in use. The germ lines in the access network 10 are prone to faults. The main causes of these faults are the entry of water and the physical damage of the nodes through which the terminal lines pass between the local switch 10 and the terminal equipment. There are five major faults that occur due to the causes that arise in the nodes. These failures are disconnection, short circuit, faulty battery voltage, ground leakage and low insulation resistance. A disconnection arises when a terminal line is interrupted between the local switch and the terminal equipment. A short circuit arises from wire A and wire B from a line that are connected together. A faulty battery voltage arises when wire A or wire B of a terminal line has a short circuit connection to one of the wires of another line. A ground leakage arises when wire A or wire B is connected to ground. The resistance to low insulation arises when the resistance between wire A and wire B or between one of the wires and the ground or between one of the wires and a wire of another line is below an acceptable value. To detect faults in the terminal lines of the access network 12, the local switch 10 is provided with a line tester 80. The line tester 80 can be operated from the local switch 10 or as will be explained in more detail below, from a remote location. The line tester 80 is capable of running several tests, examples of which will be described below. Several models of line testers for local switches are commercially available. In the present example, line tester 80 is supplied by Porta Systems of Coventry, England. . Referring now to Figure 2, there are some of the switches of the telecommunications network in which the local switch 10 is located. In addition to the local switch 10, Figure 2 shows two main switches 90, 91 and an additional local switch 92. The main switches 90, 91 are part of a fully interconnected network of main switches. The local switches 10 and 92 are part of a much larger number of local switches. Each local switch is connected to one or two main switches. Therefore, the main switches 90, 91 connect the local switches together. Referring now to Figure 3, the local switch 10 and telecommunication network components that provide a fault management sm for the access network 12 are shown. These components comprise the line tester 80, a service sm of clients 100 for the telecommunications network and an access network management sm 102. As shown in Figure 3, the line tester 80 comprises a hydrostatic test height 104 that contains the electronic equipment for physically performing line tests and a controller 106 for the test hydrostatic head 104. The controller 106 takes the form of a controller. The controller 106 can be operated from a workstation 108 connected to it and provided in the local terminal 10. The controller 106 is connected to the customer service sm 100 and the access network management sm 10 * 2 and can be operated by work stations connected to either the customer service sm 100 or the access network management sm 102. The client sm 100 is also a computer and can be operated from a number of workstations, which are connected to the. In Figure 3, one such workstation is shown and indicated by the reference number 110. The customer service sm 100 is used by operators of the public telecommunications network that has direct telephone contact with customers of network. Together with these operators, the customer service sm is responsible for providing various services to customers. These services include the provision of new telephone lines, answering the billing questions and responding to the failure reports received from customers. The access network management sm 102 is also a computer and can be operated from a number of work stations. One of these work stations is shown in Figure 3 and is indicated by the reference number 112. The access network management sm 102 is responsible for the access network 12 as well as other access networks in the same. general geographic area such as the access network 12. The access network management sm manages several operations for each of the access networks it manages. These operations include the provision of new equipment, recording the data on the work performed by the engineers in the network, maintaining the data on the terminal lines and nodes of each access network detection and fault management. The workstations that are connected to the access network management sm 102 are also connected to the customer service 100. As shown in Figure 3, the customer service sm 100 and the access network sm 102 are connected. together. The operations executed by the customer service sm 100 and the management sm 102 other than the stopping and handling of the failures of the access network 12 do not form part of the present invention and will not be described in greater detail. Although the present example of the fault management sm for the access network 12 is formed from the line tester 80, the customer service sm 100, the access network sm 102, the fault management sm could be provided simply by using the line tester 80 itself. To achieve this, it would be necessary to add the appropriate computer programs to the computer that form the controller 106. In a small network, this would be a way to provide the fault management system. However, in a large public network it is advantageous to even integrate the fault system management into the customer service system 100 and the access network system 102.

As mentioned before, the controller 106, the customer service system 100 and the access network management system 102 are each implemented as a computer. The following main electromechanical components of a computer are shown in Figure 4, these comprising a central processor unit (CPU) 120, a storage 122, a keyboard 124, a visual display device (VDU) 126 and input and output ports. 128 to connect the computer to other computers. The storage 122 comprises hard disk storage, floppy disk storage, random access memory and read only memory. The storage 122 is used to store the data used by the computer and also to store the programs that control the computer. As will be explained below, in the failover management system, one of the parameters used to find a fault in a terminal line, as long as it is available, is the distance of each node through which it passes. terminal line from local switch 10. In a typical access network those distances will be known by some, but not for all nodes. For some of these nodes, the distance is known from maps of the access network cables. In the present example, when those distances are known from maps, the appropriate data is maintained both in the customer service system 100 and in the access network management system 102. The failure management system of this example it also provides a method for measuring the distance of a node from the switch 10, when the node is opened by the engineer. The flow diagram of the operations that are executed when the distance is not measured, are shown in Figure 5 and those operations will now be described. Initially, in a step 140, an engineer opens a node and makes a telephone call to a colleague who is in one of the work stations connected to the access network management system 102 to ask him, if he performs line tests. Then, disconnect one. of the node line pairs and gives the directory number of this line pair to its colleague in a step 142. Using the connection between its workstation and the line tester 80, the colleague instructs for the line tester to 80 execute the line tests on the open line pair. In those line tests, the resistance between wire A and wire B, the capacitance between wire A and wire B, the capacitance between wire A and ground and the capacitance between wire B and ground are measured in one stage 144. Then, in a step 146, the line tester 80 uses the results of the tests to calculate the distance from the switch to the node. Next, in a step 148 the result of the tests including the distance to the node are stored in the customer service system unless the node is a "union." Finally, in a step 150, the results of the tests which include the distances and the node, are stored in the access network management system 150. Those results are stored in step 150 for all the nodes including the junctions When executing the operation shown in the flow diagram of the Figure 5 each time a node is opened, the reference data including the distances of the nodes from the switch can be gradually accommodated The series of operations shown in Figure 5 are executed under the control of computer programs located in the controller 106 , the customer service system 100 and the access network management system 102. The controller 106 is programmed to cause the hydrostatic head of test 104 to make a series of tests. You were routine every night on each access terminal line 12. These tests will be explained with reference to the circuit diagram shown in Figure 8. To test a line, disconnected from the tester 10 and connected to the test hydrostatic height 104. Figure 8 shows a line 400 being tested. The line 4 00 has a wire A 402 and a wire B 404. The end of the line 400 remote from the switch 10 is connected to the terminal equipment 406. Each of the lines 402, 404 have a resistance that depends on its diameter and the distance from the local computer to the terminal equipment 406. Each of the cables 402, 404 is coated with an insulating material, which may be a plastic or paper material. The function of the insulating material is to provide the insulation between each cable and the adjacent cables. Damage to the insulating material or oxidation of the metal of an insulating material can cause the resistance between the two adjacent wires to fail. The effectiveness of the insulation between the wires 402, 404 can be determined by measuring the resistance Rl between the wire A 402 and the wire B 404 and the resistance R2 between the wire, B -404 and the wire A 402. The resistors Rl and R2 they may be different due to the rectification as indicated by diodes DI and D2. For a circuit in good condition, the resistors Rl and R2 are high, greater than one megaohm. Damage to the insulating material or oxidation will cause the resistors Rl, R2 to fail by an amount that depends on the severity of the damage or oxidation. If the insulating material is completely destroyed so that the wires A and B are physically touching each other, the values of the resistance Rl and R2 will depend on the distance between the test hydrostatic head 80 and the point of damage although they will typically be located in the scale from 0 to 1500 ohms. Oxidation can result in the wires effectively touching each other. Only wires A and B 402, 404 of line 400 that are tested are disconnected. In the other lines, the 50 volt deviation voltage is applied between the wire A and the wire B. In Figure 8, the wires A of the other lines are collectively shown by a wire 410 which is connected in the switch 10 to Earth. The wires B of the other lines are shown collectively by a wire 412 connected to the switch to a potential of 50 volts. If the insulating material that separates material A 402 or wire B 404 from an adjacent one of wires A or B is damaged, or if one of the wires undergoes oxidation, the current can flow. The effectiveness of the insulation between the wires A and B 402, 404 and the adjacent wires A and B can be determined by measuring the resistance R3 between the wire A 402 and the adjacent wires A 410, the resistance R4 between the wire A 402 and the adjacent wires B 412, the resistance R5 between the wire B 404 and the adjacent wires A 410, and the resistance R6 between the wires B 404 and the adjacent wires B 412. For a good circuit, the resistors R3, R4, R5, R6 are high , greater than one megaohm. Damage to the insulating material can cause one or more resistors R3, R4, R5, R6 to fail in an amount that will depend on the severity of the damage. If the insulating material between the A 402 wire or the B 404 wire and an adjacent wire is completely destroyed so that the two wires are physically touching each other, the resistance between the two wires touching each other will depend on the distance between the hydrostatic head of test 80 and the point of damage although it is typically located on the scale of 0 to 1500 ohms. Oxidation can result in two different wires that touch each other effectively. The wires A and B 402, 404 and the insulating material between them act as a capacitor. In Figure 8, the capacitance between wires A and B is shown having a value of * Cl. The value of the capacitance between wires A and B of a line will depend on the length of the line. A separation at line 400 will reduce the capacitance Cl value as measured from test hydrostatic head 80. Figure 8 also shows capacitance C2 between wire A 402 and capacitance C3 between wire B 404 and ground . Each night, the controller 106 causes the hydrostatic test height 80 to measure the resistors R1, R2, R3, R4, R5, R6 and the capacitances Cl, C2, C3 for each terminal line of the access network 12. The access controller 106 also causes the hydrostatic test height 80 to verify whether terminal equipment is connected to the. end of the line. The terminal equipment has a standard capacitance value. When the terminal equipment is connected, the value of its capacitance is subtracted from the capacitance as measured by the hydrostatic test height to obtain the capacitance Cl. For each terminal line, the test results are stored against its directory number in the access network management system 102. The controller 106 transmits the test results of the access network management system * 102. The access network management system 102 examines the results of the test series of each terminal line for the presence of a supposed fault. Possible faults include disconnection, short circuit, faulty battery voltage, earth leakage and low insulation resistance. When a failure is suspected, the name of the fault and the results of the test for the line are stored in the access network management system 102 against its directory number or an identifier in the terminal associated with the line. The details of the alleged failures found each night can be reviewed by an operator of the access network system 102. When appropriate, the operator can give instructions for a fault to be repaired. . As mentioned before, the data referring to the terminal lines and the nodes are stored in the access network system 102. For each node, these data include the history of the engineering interventions in the node, the presence or absence of an active engineering intervention in the node and the presence or absence of an active adverse visual support on the condition of the node The data on the loss of pressure in the main cable 14 are also stored in the network of the management system access network 102. A map of the access network 12 is stored in the system, customer service 100 and the access network management system 102. For each line and therefore for each circuit, the map records the route and therefore each node through which the line passes, for each node registers the lines that pass through the node.

Referring now to Figure 6, there is shown a flow chart of the series of operations that can be executed when a client reports a failure on one of the terminal lines of the access network 12 to an operator of the network of the customer service 100. As will become evident, in an appropriate case a list of the nodes through which the terminal line passes is sorted in order of probability that the failure is present in each node. Initially a client reports a failure on one of the terminal lines to an operator in the customer service system in a step 200. If the operator considers the fault to be in the access network 12, the operator sends an instruction to the controller 106 of the line tester 80 to execute a series of tests on the terminal line. The line tester 80 then executes a series of tests that it executes during the night tests on the terminal lines. The results of the test series are transmitted by the line tester 80 to an access network system 102. The access network system 102 examines the results for the presence of an assumed fault and, when appropriate, calculates the distance of the failure from the local switch 10. The details of the alleged failure are transmitted to the customer service system 100. When the operator in the customer service system considers that it is appropriate-that the-failure-be investigated by a operator in the access network management system 102, in a step 201, the operator in the customer service system asks the operator in the access network system 102 to investigate the failure. The operator in the access management system then decides what to do about the failure report. This decision will depend on the details of the alleged failure and its knowledge of the status of the terminal lines in the access network 12. For example, if the failure report indicates that there is a disconnection and the operator knows that one of the nodes through of which passes a terminal line ha. If it has been damaged severely even though it is under repair, it may be appropriate for the operator to advise the customer only that the damage is being repaired. Usually, the operator will decide to obtain a list of the nodes through which the terminal line passes for the probability that the failure is present in each node. If the operator decides to obtain this list, in a step 203, the operator invokes a classification algorithm. The rest of the flow diagram shown in Figure 6 refers to the steps that are executed in this algorithm. The classification algorithm is accessed in a step 204. The classification algorithm is maintained as part of a program of the access network management system 102. As will become apparent from the following description, for each of the nodes in the terminal line over which a failure is suspected, an individual record is evaluated for each of U? set of factors related to the probability that an assumed fault is present in the node and the individual records are combined to produce a combined record for the node. The nodes are classified after according to their combined records. In the evaluation of each individual record, the data contained in the report, of failure and also the data contained in the access network system are related to the nodes and the terminal lines are used.

After accessing the classification algorithm, in a step 205, the route followed by the terminal line over which a failure is suspected is identified and all the nodes on this route are found later. Next, in a step 206, the first node on this rut-a is selected. In a step 207, the individual record for the first factor is evaluated. The first factor is the analysis of historical intervention. The registration for this factor is evaluated in five stages as follows: stage 1: for at least one engineering intervention in the previous five days, registration 3000; stage 2: for at least one engineering intervention in a period of 30 days prior but excluding the period of five previous days, registration 2000; stage 3: for at least one engineering intervention in a period of 90 days before but excluding the previous 30-day period, record 1000; Stage 4: for the total number of engineering interventions in the previous 90-day period, it records 100 times the number of interventions. The total record for the historical intervention analysis factor evaluated in step 207 is then the sum of the records found in stages 1 to 4 above. - - - •. ~ "The next factor is the analysis, of visual inspection and it is evaluated in stage 208. If there is an adverse active visual report for the node, the record for this factor is evaluated in stage 208 as 2000. The next factor is the active intervention analysis and this factor is evaluated in stage 209. This factor is evaluated in two stages.In the first stage, if there is an active engineering work in the node, a record of 3000 is made. In the second stage, if there is an active engineering job for any terminal line that passes through the node, a record of 3000 is made. The total record for this factor is the highest of the records found in stages 1 and 2. The next factor is the assumed line analysis and this factor is evaluated in step 210.

For each terminal line that passes through the node and over which a failure is suspected and which is one of the same type as the assumed failure under a study, a record of 1000 is made. Therefore, the assumed failure is a disconnection , and the disconnection is also assumed on three of the other terminal lines that pass through the node, then the total register is 3000. Then, in a step 211, a verification of the presence of the registration factors is made between the visual inspection analysis factor as evaluated in step 208, the active intervention analysis factor as evaluated in step 209 and the assumed line analysis factor as evaluated in step 210. If two or three of those factors are recorded, then the combined record of the factors evaluated from steps 208 to 210 is the sum of the individual factors multiplied by two. If only one of these factors is recorded, then the total record of the factors evaluated in steps 208 to 210 is the record for that individual registration factor. Next, in a step 212, the record is evaluated for the proximity analysis factor. The registration for this factor is evaluated in two stages. As will be recalled from the previous description, the access network management system contains data on the distance between the local switch 10 and the new derivatives from two sources. The first source is the measurements that are made as a result of the opening of the nodes engineers and execute a series of operations shown in the flow diagram of Figure 5. The second source of information are the distances that are found from the maps of the terminal lines. The proximity analysis is executed using the data sources. For the first source of data, which are the distances derived from the opening of the nodes, the record is evaluated as follows: if the proximity of the node to the calculated location of the assumed fault is less than 50m. record 3000; if the proximity of the node to the calculated location of the assumed fault is greater than or equal to 50m. but less than 100m, record 2000, if the proximity of the node to the calculated location of the fault is equal to or greater than 100m and less than 150, register 1000, for the second data source, ie the data derived from the maps of the terminal lines, the record is evaluated in the same way as specified for the first data source. The total record for the proximity analysis factor is then obtained by adding the records obtained and using the data from the two sources. In stage 212, when it is. evaluates the record for each data source, if the distance is not available for the distance of the node from the switch to that data source, then the record is 0. Then, in a step 213, the record is evaluated for the factor of pressure analysis. The pressure is the pressure from the cable 14 leading from the switch 10 to the primary cross connection point 16. If the pressure is below a threshold value, the register is evaluated at 3000. Next, in a step 214 , the operator is invited to add a manual record, and normally, an operad.or will do this if he knows something about the special problem. Next, in a step 215, each registration factor is compared to the type of fault and an additional record is made according to Table 1 set forth below. In this Table, the visual inspection analysis factor of step 208, the active intervention analysis factor of step 209 and the assumed line analysis factor 210 are treated as a single factor. The proximity analysis factor evaluated in step 212 is divided into two subfactors _. The first subfactor is the proximity analysis performed using the first data source, that is, the data obtained through the node opening and the calculation of the distance from the line tests. The second subfactor is the proximity analysis factor evaluated from the second data source, that is, the maps of the terminal lines. Referring to Table 1, for example, if the result is a short circuit, additional records are made as follows. If there is a record for the historical active intervention factor of step 207, an additional record of 3000 is made. If there is a record for one or more of the factors of, steps 208 to 210, an additional record of 1000 is made If there is a record for this pressure, the analysis factor of step 213, an additional record of 4000 is made. If there is a record of analysis factor derived from the first data source, a record of 4000 is made. there is a record of proximity analysis factor obtained from the second data source, a record of 4000 is made. The total additional weight record is the sum of the individual records.

TABLE 1 Additional Weight of Type Defalla Then, in a step 216, a verification of the coincidence in the registration factors is made. More, t. -specifically, if there is only one individual registration factor, a jump is made to step 218. If there are two or more registration factors, the algorithm goes to step 217 before proceeding to step 218. For the purposes of In step 216, the factors of steps 208 to 210 are combined and the factor of step 212 is divided into subfactors in the manner discussed in relation to Table 1 above. In step 217, if there are two or more registration factors, the highest record is taken and increased by 100O0 for this purpose. For steps 208 to 211, the record obtained for those stages at the end of 211 is used. The record obtained in that way then becomes to combine the record of the individual factors for the node. If there is only one registration factor, this factor becomes the combined record for the nodes. In step 218, a "check is made to determine if there are more nodes for which a combined record is evaluated. If one or more remaining nodes still exist, the algorithm returns to step 206 and the next node is selected. If there are no more nodes, the algorithm continues with step 219. In step 219, the nodes are classified for registration in descending order. Therefore, the nodes at the top of the list are the most likely failure locations. In step 210, the node with the highest register is selected later and in step 211 the operator receives the opportunity to omit the result and the operator will normally do this in view of the special circumstances known to him. Referring now to Figure 7, a functional block diagram of the failure management system described with reference in Figures 3 to 6 is shown. As shown in Figure 7, the failure management system comprises the hydrostatic head test 104, the hydrostatic test height controller 106, a test report analyzer 300, a storage 301 and a node identifier 302. The test hydrostatic head 104 and the test hydrostatic head 106 have already been described. The test report analyzer 300, the storage 301 and the node identifier 302 are implemented by the access network management system 102. More specifically, the test report analyzer is the part of the access network management program which is responsible for executing steps 206 to 221 of the classification algorithm. The storage 301 is that part of the computer storage used to implement the access network management system 102 and which contains the data related to the nodes and the terminal lines. The node identifier 302 is that part of the program that is responsible for step 205 of the classi fi cation algorithm. They will now be described with reference to the Figure 9 the individual stages of one. routine that is executed every night by the access network management system 102 to monitor the operating condition of the individual nodes of the access network 12. As will become apparent from the following description, the routine uses the resistance values Rl to R6 for each circuit obtained for the night test routine in the individual circuits. The routine shown in Figure 9 also uses the map of the access network 12 which is stored in the access network management system 102. Referring now to Figure 9, after accessing the routine, in a step 500 a register circuit \, S is calculated for each circuit from the resistance measurements Rl to R6. The circuit S is indicative of both the probability that a circuit has or develops a fault and the operational quality of the circuit. A relatively high circuit register indicates that there is a high probability that the circuit has or develops a fault and that the operational quality of the circuit is poor. A relatively low circuit register indicates that there is little likelihood that the circuit has or develops a fault and that the operational quality of the circuit is good. To calculate the circuit register S for a circuit, each resistance value Ri is converted into a converted value Vi using the formula that will be described below. The converted value Vi is indicative of the probability that the resistance will cause a failure. Each converted Vi value is multiplied by a Wi factor to obtain the product Vi * Wi. The Vi * Wi products for the six resistance measurements are summed to produce the S circuit register. Therefore, the circuit register of S is defined by the following equation: er \ - S = SVi * Wi Each converted value Vi is calculated for the corresponding resistance value using the following formula: If Ri = P, Vi = O If P < Ri = Q, Vi = 1 If Ri > M, Vi = O If Q < Ri = M then . { [1 / (1+ ((Ri-Q / L) 0-3)] + [-! * (Ri-l * 105) / (1 * 106-Q. /. / 2 where P is a fixed lower threshold kO Q is an upper threshold set at 5 kO M = 1MO L is constant set at 1 kO A graph of Vi plotted against Ri using the formula established above is shown in Figure 10. The formula for calculating Vi has been developed empirically. However, the physical explanation behind the formula is as follows: When a resistance Ri of a circuit has a value less than lkO, it is very likely that the cause of the low resistance is a fault in the terminal equipment, and so both the value Vi is set to zero. When a resistance Ri has a value on the scale of lkO to 5kO it is almost certain that there is an associated fault. For a resistance value greater than 5kO, the probability of existence of a fault associated with a failure progressively and less progressively staggered. .. with the . increase value of Ri. When a resistance Ri has a value greater than 1MO, it is less likely that there is an associated fault. For each type of resistance Rl to R6, the value of the corresponding weight factor is determined from the following table: Types of resistance Weight factor Ri Rl (wire A to wire B) 15 R2 (wire B to wire A) 15 R3 (wire A to ground) 5 R4 (wire A to -50V) 30 R5 (wire B to ground) 30 R6 (wire B to -5pV) 5 After determining the circuit register S for each circuit is stage 501, the routine proceeds to step 502. In this step, it calculates a register of node H for each node.

To calculate the node record H for a node, all circuits passing through the node are identified. The individual circuit registers S for the individual circuits passing through the node are added afterwards and the result is "divided by" * / n. n is the number of resistance measurements for circuits that pass through the node that has a value less than 1MO. Therefore, the node register H is defined by the following equation: The node record H for a node is indicative of the probability * of the operating state of the node that causes a failure in one or more of the circuits passing through the node and of the operational quality of the node. A relatively high node register indicates that there is a high probability that the operating state of the node will cause a failure of the circuit and that the operational quality of the node is poor. A relatively low node register indicates that there is a lower probability that the operating state of the node will cause a failure and that the operational quality of the node is good.

In the equation set before to calculate the register, the node H.? S is divided by n to provide an average that affects the resistance measurements in the circuits that pass through the node that has a value less than 1MO. This makes the node records for the nodes that carry a large number of circuits comparable with the node record for the nodes that carry a smaller number of circuits. As explained below, the nodes are classified according to a node record. However, node registers of nodes that carry only a few circuits are very sensitive to the number of resistance measurements less than 1MO. More generally, as the number of circuits passing through a node increases, and therefore potentially the number of resistance measurements less than 1MO increases, the reliability of the node record as an indicator of the operational quality of the node is increased. In the equation before established to calculate the register of node H, SS / n is multiplied by Vn to give progressive emphasis to the register of node according to the number of resistance measurements less than 1MO is increased.

Next, in a step 502, the nodes are classified according to their node H registers. The node having the highest node register H is identified as the worst node and selected for further investigation. Although the relatively high node register of the worst node can be caused by the operating state of the worst node, it can also be caused by the operating state of one of the nodes in the route from a local switch to the worst of the nodes. The record of node H is obtained from the resistance measurements and does not take other factors into account. To identify the precise location of the node that has * poor operational quality, the classification algorithm described with reference to Figure 6 is used. Accordingly, in step 503, a route from the local switch 10 to the worst node is found and all the nodes in the route and including the worst node are found. Steps 206 to 219 of the classification algorithms are executed by this route in a step 504. Occasionally, there is more than one route from a local switch to a node. In a step 505, a check is made for the existence of another route to the worst of the nodes. If one or more additional routes exist, steps 505 and 504 are executed for all additional routes. Finally, in a step 506, the operator decides whether to ask an engineer to investigate one or more of the nodes for possible failures. When the routine reaches step 506, the operator knows the identity of the worst node. Also, for the nodes in each route to the worst node, you will have a list of the nodes as they were classified using the classification algorithm. In deciding whether to ask the engineer to investigate one or more of the nodes, the operator combines this data with his own knowledge of the access network 12. In addition to investigating the worst node, the operator can investigate other nodes, example the next worst node or other nodes that have high node records. In order to do this, steps 504 through 506 are executed for each of those nodes. , t. • The routine of Figure 9 makes it possible to identify and investigate at an early stage the nodes when there is a risk that the operational state of the node is deteriorated to a point where failures may occur. It will usually be possible to restore the operating status of such a node before a failure occurs that leads to a failure report. Figure 11 shows the results of some experimental work on the nodes in an access network that is part of the. public telecommunications network of the United Kingdom BT 's. The node record was evaluated for a large number of nodes. Each node was then monitored for a failure report from a client during the subsequent three months. In Figure 11, for those nodes, the node record is plotted against the failure reports received during the three months after the evaluation of the node record. These experimental results show a strong correlation that considered the node record and the number of fault reports. Although the present invention has been described with reference to an access network in which each circuit is transported by a piece of copper wire, it can be used. also for terminal circuits transported by optical fibers.

Claims (15)

1. A fault management system for a telecommunications network that includes a local switch and a set of terminal circuits that extend between the local switch and. the terminal equipment provided for the users of the network, each of the terminal circuits passing through a series of nodes between the local switch and its respective terminal equipment, the fault management system comprising: the test apparatus of circuit located in said local switch and placed to execute circuit tests in the terminal circuits; a storage that contains data related to terminal circuits and nodes; means for instructing the circuit test apparatus to execute a series of tests on one of the terminal circuits; means for verifying the results of a set of tests executed by the circuit test apparatus for the presence of a supposed fault, the means of verification that are placed to produce a fault report when a failure is suspected; means for identifying the nodes of a terminal circuit over which an assumed fault is present; means to evaluate a record of each node on a terminal circuit line in which a failure is suspected that represents the probability that the alleged failure is present in the node, the means of evaluation, using the failure report that relates the alleged failure and the data contained in the data storage in the evaluation of the record; and means for classifying nodes of a terminal circuit over which a failure is suspected in accordance with its registers encentrados by the evaluating means; whereby, in use, after a test of a circuit on which a failure is suspected, a list of the nodes in said circuit is produced in which the nodes are classified according to the probability that the failure is present in each node.

2. A fault management system as claimed in claim 1, wherein for each node in a terminal circuit over which a failure is suspected, the evaluation means are set to evaluate an individual record for each of a set of factors and to combine the individual records to produce a combined record representative of the probability that the assumed failure is present in the node. ~~

3. A fault management system as claimed in claim 2, in which the set of factors includes the history of the engineering interventions in the node.

4. A fault management system as claimed in Claim 2 or Claim 3, in which the set of factors includes the presence or absence of active engineering intervention in the node.

5. A fault management system as claimed in any of claims 2 to 4, in which the set of factors includes the presence or absence of a current adverse visual report on the node.

6. A fault management system as claimed in any of claims 2 to 5, in which the set of factors includes the presence or absence of a suspected failure in another terminal circuit that passes through the node.

7. A fault management system as claimed in any of claims 2 to 6, wherein the fault report includes an estimate of the location of the fault and the set of factors includes the proximity of the calculated location of the fault to the location of the node.

8. A fault management system for a telecommunications network, the telecommunications network that includes a local switch and a set of terminal circuits that extend between the local switch and the terminal equipment provided for the user of the network, each of the terminal circuits passing through a series of nodes between the local switch and its respective terminal equipment, the fault management system comprising: the circuit test apparatus located in the local switch and placed to perform circuit tests on the terminal circuits; a computer system for operating the circuit test apparatus; ~ the computer system that includes a storage containing data related to the terminal circuits and the nodes; The computer system that is controlled by at least one program to execute the following operations: instruct the circuit test apparatus to run a test set on one of the terminal circuits; verifying the results of a set of tests performed by the circuit test apparatus for the presence of an assumed fault and producing a fault report when the fault is suspected; identify, .. the nodes of a terminal circuit over which a supposed fault is present; evaluate a record for each node on a terminal circuit over which a fault is assumed which represents the probability that the assumed fault is present in the node using the failure report in relation to the assumed failure and the data contained in the storage; and classifies the nodes of a terminal circuit in which a failure is suspected as claimed with its records.

9. A method of operation in a fault management system for a telecommunications network, the telecommunications network that includes a local switch and a set of terminal circuits that extend between the local switch and the terminal equipment provided for the users of the network , each of the terminal circuits passing through a series of nodes between the local switch and its respective terminal equipment, the fault management system comprising: apparatus for circuit testing located at the local switch, and - positioned for execute circuit tests on the terminal circuits; and a computer system for controlling the circuit test apparatus, the computer system that includes a storage containing data related to the terminal circuits and the nodes; the method comprising the following steps executed by the computer system: instructing the circuit test apparatus to execute a test suite on one of the terminal circuits; verify the results of the set of tests performed by the circuit test apparatus for the presence of an assumed fault and produce a fault report when the assumed fault is present; when a fault is assumed, identify the terminal circuit nodes on which the failure is suspected; evaluate a record for each of the nodes of the terminal circuit over which a failure is suspected, which represents the probability that the assumed failure is present in the node based on the report of / a. ia that is related to the supposed failure and using the data contained in the storage of data; and classifies the nodes of the terminal line on which the failure is suspected according to the records found in the record evaluation stage.

10. A method of operation in a fault management system as claimed in claim 9, wherein, in the step of evaluating a record for each of the nodes, for each of the nodes an individual record is evaluated for each one of a set of factors and the individual records are combined to produce a combined record representative of the probability of the assumed failure that is present in the node.

11. A method of operation of a fault management system as claimed in claim 10, wherein the set of factors includes the history of engineering interventions of the node.

12. A method of operation of a fault management system as claimed in claim 10 or claim 11, wherein the set of factors includes the presence or absence of the active engineering intervention in the node.

13. The method of operation of a fault management system as claimed in any of claims 10 to 12, wherein the set of factors includes the presence or absence of a current adverse visual report on the node.

14. A method of operation of a fault system as claimed in any of claims 10 to 13, wherein the set of factors includes the presence or absence of an assumed failure on another terminal circuit passing through the node.

15. A method of operation of a fault system as claimed in any of claims 10 to 14, wherein the fault report includes a estimation of the location of the fault and the set of factors includes the proximity of the location calculated from the fault towards the location of the node. SUMMARY An operation method is described in a fault management system for an access network that is part of a public telecommunications network. The access network, the terminal lines in the form of pairs of copper wires, is They extend from a local switch (10) through a series of nodes to the terminal equipment provided to the network user.The fault management system includes a hydrostatic test height (104) and a network management system of access (102) .Every night, the hydrostatic test height (104) executes a series of tests "on each of the terminal lines.The results of the test are transmitted to the access network management (102) .The test results are then converted into registers circuit, each of which is indicative of the operational quality of the circuit tested.For each, no-do, the circuit logs of the circuits tested passing through the node are combined to produce a node record that is indicative of the operational quality of the node To identify the node that has the worst operational quality and therefore that has the greatest need for research. the nodes are classified according to their node records.