COMMUNICATIONS CIRCUIT TEST The invention relates to testing in communications networks.

Modern networks increasingly share a physical path between two or more services. In the access network, for example, more and more subscriber PSTN lines are being upgraded to support broadband DSL traffic in addition to the legacy voice traffic. The subscriber line may be implemented in a number of technologies, normally copper or optical fibre but other suitable technologies, such as radio frequency links may also be used. There is a need to verify the correct configuration of services in the access network and, in particular, to check the correct combination of services on a specific access line. To achieve this it will be necessary to identify both errors in the wiring in the access network and incorrect records in the databases recording aspects of the configuration of the access network. The two scenarios where verification is important are the transferring of an existing service (transfer engineering) and the provisioning of a new service. In the case of provisioning a new service, the source of a fault may lie in either the new wiring or the updated database, although modern systems with automatic loading of databases reduce the chances of database error. In the case of transfer engineering, the physical connections may have been in place for some time so that any fault is likely to relate to a service change (e.g., moving houses, ceased or upgraded services) that are likely to cause mismatching data recorded in various different databases.

One section of the access network prone to connection errors is the Main Distribution Frame (MDF). The MDF is located at a telephone exchange or central office and is divided into two sides. On the D-side, the MDF provides terminations for cables connecting to telecommunications terminal equipment, (e.g. telephone apparatus) located at users premises. (e.g. telephone subscribers). On the exchange or E-side, the MDF provides terminations for cables for connection to the telephone switch and other: communication company equipment such as DSLAMs - depending on whether it is a broadband DSL or narrowband POTS circuit. Users are connected to the appropriate port on the switch or DSLAM by manually connecting a jumper wire between the user connection on the D-side and the appropriate connection on the E-side of the MDF.

The manual wiring involved in setting up and maintaining a MDF inevitably leads to some incorrect connections being made. Data on the planned configuration of a MDF is

stored in database-form to allow maintenance and upgrade work to be planned by the network provider. This configuration data is "planned" in that it reflects the desired configuration but this may not correspond to the actual, physical configuration if incorrect connections have been made. Indeed, the act of manually recording the jumper settings will tend to lead to database errors, so that the wiring present in the MDF is not accurately recorded. There is therefore a need to verify the data in the MDF database and reconcile it with the reality of the physical MDF (an activity known as "data cleanse") to allow inappropriate changes in the configuration to be identified and corrected. The importance of having accurate records is emphasised if a major overhaul of the access network is planned. In addition to the MDF data, service data is also recorded in the database, e.g., DNs, PSTN and DSLAM port numbers.

In practice, rather than a single database, in the BT system, there are two separate databases: POTS and DSL whose contents are related as they deal with different services of the same access network. The content of these databases needs to be compared to provide a further check on the accuracy of the data.

Examples Of Mismatching Parameters

The design of test and diagnostic solutions for broadband data verification is based on a detailed understanding of the scenarios in which data mismatch problems can occur. Various scenarios are set out in Table 1, below. In general, a DN and a DSLAM port are associated in one database to form a pair but, in the other database, either the DN is associated with a different DSLAM port or the DSLAM port is associated with a different DN, giving rise to mismatched pairs.

In the first scenario "Mismatch scenario-1" the DN of a broadband line is re-numbered on the POTS database (i.e. to DN2) but the DSL database is not updated accordingly (e.g. due to an error in the manual process for updating the DSL database). This mismatch case is shown in the second row of Table 1.

The second scenario "Mismatch scenario-2" occurs when the DSLAM port (e.g. port 1) of a broadband line (associated with DN1) is changed for another port (e.g. port 2) and the DSL database is updated with a new network ID (i.e., DSLAM port number) while the POTS database is not updated accordingly. This leads to the mismatch shown in the third row of Table 1.

Other scenarios include "Mismatch scenario-3": the combination of the first two scenarios; and "Mismatch scenario - 4": the mistyping of DNs or DSLAM port identifiers as shown in the last row of Table 1. Other mismatch scenarios that may occur are addressed by the present invention but are not treated here for reasons of brevity.

Table 1. Scenarios of broadband data verification

A conventional test for access lines known, as the OFJ test, relies on detecting the capacitance added to the line by the presence of a DSLAM. A problem with this is conventional test is that some second-generation DSLAMs add very little capacitance to the line. As a result, the OFJ Test produces a high percentage of inconclusive results. There is therefore a need for a more reliable test regime for access networks.

The present invention provides a method for verification of a set of access circuit parameters, in which the parameters indicate an access circuit; and an equipment port on an item of equipmentr-the method including the steps of

(a) testing the access circuit;

(b) modifying the characteristics of a line associated with the equipment port and

(c) testing again the access circuit tested in step (a) so as to determine if the equipment port is connected in the access circuit. The step of modifying the characteristics of the line may involve changing the complex impedance of the line at the equipment port or disconnecting an access circuit at a point between a user and the equipment port. Changing the complex impedance of the line may involve one of removing a reactive component from the line at the equipment port and adding a reactive component to the line at the equipment port.

According to an aspect of the invention, a first database is provided for access circuit information and a second database is provided for equipment port information, in which the method includes inputting user information relating to one of the access circuit and the equipment port and obtaining the other of the access circuit and the equipment port information by database lookup.

According to a further aspect of the invention, the first database is provided for associating a set of access circuits with a set of broadband network identifiers and the second database is provided for associating a set of equipment port with a set of broadband network identifiers, in which the method includes forming an association between an access circuit and an equipment port associated with the same broadband network identifier.

The method preferably includes determining that information on a first access circuit is missing from the first database and substituting for the purposes of the test information on a second access circuit adjacent to the first access circuit.

The method preferably includes determining that information on a first equipment port is missing from the second database and substituting for the purposes of the disconnection information on a second equipment port adjacent to the first equipment port.

According to a further aspect of the invention, the method includes verifying two pairs of parameters, in which each pair comprises access circuit information and equipment port information; in which the method includes applying the method of any above claim to each of the two pairs and interpreting the test results to determine the correctness of

each pair. Preferably, the two pairs comprise information on the same equipment port and information on different access circuits, or the two pairs comprise information on the same access circuit and information on different equipment ports.

According to a further aspect of the invention, the first database is a narrowband service-related database and the second database is a broadband service-related database. According to further aspects of the invention, the access circuit is indicated by a user DN; the item of equipment is a multiplexer; the access circuit is associated in the set of parameters with the equipment port.

The invention also provides a tester for circuits, each circuit for carrying an instance of a first and an instance of a second communications service, each service comprising a plurality of instances and each instance being associated with one of the circuits; in which the tester comprises means for performing a first and a second test on the circuit associated with a first instance of the first service; means for applying a condition to the circuit associated with a first instance of the second service during the second test but not during the first test; means for analysing the results of the first and second tests to detect in the circuit associated with the first instance of the first service the condition applied during the second test.

According to a further aspect of the invention, tester comprises means to perform a third and a fourth test on the circuit associated with the first instance of the first service; in which the tester further comprises means for applying a condition to the circuit associated with a second instance of the second service during the fourth test but not during the third test; means for analysing the results of the third and fourth tests to detect in the circuit associated with the first instance of the first service the condition applied during the fourth test.

According to a further aspect of the invention, the tester comprises means to perform a fifth and a sixth test on the circuit associated with a second instance of the first service; ^■ in which the tester further comprises means for applying a condition to the circuit associated with the first instance of the second service during the sixth test but not during the fifth test; means for analysing the results of the fifth and sixth tests to detect in the circuit associated with the second instance of the first service the condition

applied during the sixth test. Preferably, the tester comprises means for reporting a fault if the condition is not detected.

According to a further aspect of the invention, the tester comprises a plurality of databases, one per service, in which each of the databases comprises a record of the associations between the instances of the relevant service and the circuits; in which the tester also comprises means for comparing the database entries of two services and for detecting an inconsistency between them. According to further aspects of the invention, the tester comprises means to perform the second test before the first test; means to perform the fourth test before the third test; means to perform the sixth test before the fifth test. Preferably, the condition applied during the second test is different from the condition applied during the fourth test. Preferably, the condition applied during the fourth test is different from the condition applied during the sixth test.

According to further embodiments of the invention, the condition may comprise additional resistance; additional capacitance; breaking the line.

The invention also provides a method of testing circuits in a communications network, each circuit for carrying an instance of a first and an instance of a second communications service, each service comprising a plurality of instances and each instance being associated with one of the circuits; the method including the steps of: performing a first and a second test on the circuit associated with a first instance of the first service; applying a condition to the circuit associated with a first instance of the second service during the second test but not during the first test; analysing the results of the first and second tests to detect in the circuit associated with the first instance of the first service the condition applied during the second test.

The invention also provides a method for checking for the presence of an access circuit extending from a test point and passing through a multiplexer and issuing through a port on the multiplexer via a test access matrix to a user terminal in a communications network, the method including the steps of: at the test point testing parameters of the access circuit and recording the results; operating the test access matrix to disconnect

the access circuit; testing the parameters again and comparing the results with the results from the first test to determine if the access circuit is correctly wired to the multiplexer port.

The invention also provides a method for checking for the presence of an access circuit extending from a test point and passing through a multiplexer and issuing through a port on the multiplexer to a user terminal in a communications network, the method including the steps of: at the test point testing parameters of the access circuit and recording the results; operating the multiplexer port to modify the complex impedance applied to a line passing through the multiplexer port; testing the parameters again and comparing the results with the results from the first test to determine if the access circuit is correctly connected to the multiplexer port.

Embodiments of the invention will now be described by way of example with reference to the drawings in which:

Figure 1 shows a functional block diagram of a system for implementing the invention;

Figures 2a, 2b and 3 show schematics of test paths according to the present invention;

Figure 4 shows a detail from Figure 3.

The basic concept of testing according to the present invention can be summarised as verifying associations between various network parameters by means of specifying a physical change in the network under test and seeking a detectable difference in a particular circuit. If the physical change is introduced based on the information in a first database and the check is carried out based on the information in a second database, the consistency of the information in the two databases (i.e. the network inventory) can be checked, along with the physical integrity of the network. Normally, each service will have its own database so the test is particularly relevant to situations where a physical resource is shared by more than one service. The invention involves a novel use of a test equipment matrix (TAM) usually installed in the access circuit. While the TAM is designed as a channel for connecting test equipment to parts of the access network, the present invention uses the TAM as a test stimulus for test equipment connected elsewhere in the circuit.

Some Definitions:

The Network Capacity Assignment System (NCAS) is part of the service provisioning system. Specifically, it provides capacity calculations to determine if access capacity exists in the network to support broadband connectivity, assigns broadband network capacity to customers and performs logical network configuration. NCAS includes an inventory of DSLAM equipment, ports and line cards and records services configured.

All aspects of the broadband equipment, including DSLAMs, are managed by ADSL Element Managers (ADSLEMs), such as the Fujitsu FENS-AN (see: http://www.fujitsu.com/uk/services/telecom/products/network/ ) and the Alcatel AWS (see: http://www.alcatel.com/products/productsummary.jsp?relativeP ath=/com/en/appxml/op gproduct/alcatel5523adslworkstationtcm228117971635.jhtml ). The Element Managers support the building of planned equipment, automatically discovering equipment when installed and downloading the appropriate configuration data to the network elements. The Element Managers also support capacity management activities, holding the data required to monitor capacity utilisation and passing this information to NCAS.

The Network Inventory and Spares Management system (NISM) is a database storing an association between each DSLAM identity and port number duplet with a TAM identity and port number duplet, i.e., identifying which TAM port is designated for connection in the line with which DSLAM port.

The test access matrix (TAM) is a commercially available device for providing access to a circuit such that a separate test system can test the circuit, typically local loop (i.e. access) circuits. It designed to be installed in the access circuit and includes a switch matrix for implementing disconnection and reconnection of the access line so as to connect in various items of test equipment to a section of the interrupted line. For example, with the TAM configured in a first mode, a test device may be connected to the section of the access circuit extending from the TAM to the end user. With the TAM configured in a second mode, a test device may be connected to the section of the access circuit extending from the TAM to the exchange equipment. This allows test

signals to be injected and measurements to be taken on isolated sections of the access network.

TAN and NNI relate to test equipment, similar in concept to the TAM, but integrated into the exchange equipment in the POTS line. Specifying parameters for TAN and NNI allows the exchange equipment to be configured for the line tests.

The user DN (originally "directory number") is used to represent a service identifier related to the voice band (i.e. narrowband) telephony service. In modern practice, the number of the user's account with the service provider identifies the narrowband service and any broadband service allocated to that user.

Figure 1 shows an overview of the solutions and systems involved. As shown in Figure 1 , Voice Data Migration Tool (VDMT) is connected to test controllers GTCS and GTCI via test controller GTCJ-TE. In operation, VDMT passes two pairs of unmatched DNs and CBUKs to GTCJ-TE for testing. GTCJ-TE will apply the WW test (as described below) on GTCS and GTCI. GTCJ-TE obtains raw results from GTCS and GTCI and returns to VDMT two pairs that are actually tested plus test outcomes. VDMT performs algorithms that interpret the outcomes, perform the data cleanse, or initiate manual verification as necessary. The full name for the WW test is TAM-based xDSL Frame Jumper Test'.

Two operations where these test may be of particular value are the provisioning of a new service and the transfer of existing services to a new network (i.e. BT's 21 CN transfer engineering). If a fault is identified in the first case, this could be due to faulty wiring or incorrect data. However, these databases are automatically loaded from a reliable source so that a circuit fault will be more likely to be the cause. If a fault is identified in the second case and the physical connections have been in place for some time, the fault is likely to be related to the service change.

We will now describe how the WW test sources the test parameters from the databases according to preferred embodiments of the invention. The test parameter input will specify either a DN (for test WW(DN)) or a DSLAM port number (for test WW(DSLAM port)).

According to a preferred embodiment, for a given DN, the test parameters are sourced as follows:

1 ) DN -> the POTS database (CSS) -> BB network ID (CBUK)

^• 2) BB network ID (CBUK) -> the DSL database (NCAS) -> DSLAM ID, DSLAM port number

3) DSLAM ID, DSLAM port number -> TAM database (NISM) -> TAM ID, TAM port number.

Also:

4) DN -> the POTS database (CSS) -> TAN, NNI

According to another preferred embodiment, for a given BB network ID (CBUK), the test parameters are sourced as follows:

1 ) BB network ID (CBUK) -> the DSL database (NCAS) -> DSLAM ID, DSLAM port number 2) DSLAM ID, DSLAM port number -> the POTS database (CSS) -> DN

Also:

3) DSLAM ID, DSLAM port number -> TAM database (NISM) -> TAM ID, TAM port number.

And: 4) DN -> the POTS database (CSS) -> TAN, NNI.

Missing database entries can restrict the ability to perform the tests. The above example shows the WW test focussed on a pair of access network parameters (DN, DSLAM port number) recorded as corresponding to the same subscriber line. If a specific DN, DN1 , is provided linked in the POTS database (CSS) with a broadband network identifier (CBUK2) but no entry can be found in the DSL database for a DSLAM port number linked to the same broadband network identifier (CBUK2), the WW test fails at the stage of sourcing the data from the DSL database, and the connection of DN1 cannot be tested. In this case, there will be no test results or direct evidence as to the validity of this pair. Indirect evidence as to the validity of this pair

may be obtained by testing the related pair (DN1 , DSLAM port4) obtained by substituting a DSLAM port adjacent to the port absent from the DSL database. Here "adjacent" refers to the numerical sequence of port numbers or the order of records in the database. If the test shows the original DN, DN1 , being physically connected with the substitute DSLAM port, DSLAM port4, or another DSLAM portm, then we can deduce that DN 1 is not actually connected with DSLAM port3.

In a similar fashion, if no entry can be found for DN1 in the POTS database, the WW test will also fail at the stage of sourcing the data from the POTS database, and the connection (DN1 , DSLAM port3) cannot be tested. Again, there will be no test results or direct evidence as to the validity of this pair. In this second case, indirect evidence as to the validity of this pair may be obtained by testing the related pair (DN2, DSLAM port3) obtained by substituting a DN adjacent to the DN absent from the POTS database. Here "adjacent" refers to the numerical sequence of circuit numbers or the order of records in the database. If the test shows the substitute DN2, or a further DN, DNn, being physically connected with the original DSLAM port.3, then we can infer that DN 1 is not actually connected with DSLAM port3.

Figure 2a shows first WW test path forming part of a telecommunications access network. The test path of Figure 2a extends from test head T/TH 20 to the end user (EU not shown). Starting at test head T/TH 20, the test path passed through an exchange equipment 22 exiting via POTS port 222 to which a DN is allocated. The test path follows the POTS connection for that DN between the exchange equipment 22 and MDF 24 connecting from exchange equipment port 222 to MDF E-side termination 241. The test path follows the MDF jumper 247 from E-side termination 241 to the corresponding D-side termination 244 and then the connection from this D-side termination 244 to POTS port 262 on DSLAM 26. The test path passes through DSLAM 26 exiting at the user DSL port 264 corresponding to entry POTS port 262 and continues along the connection back to MDF 24 at second E-side termination 243. The test path continues through the MDF 24 following a second jumper 248 from second E- side termination 243 to second D-side termination 242 to which the user's terminal (EU, not shown) is connected. The test path continues from second D-side termination 242 to terminate at the user's terminal in the users premises (not shown). A third

connection to DSLAM 26 connects a broadband signal between the end user's DSL line and a separate ATM network (not shown).

The exchange equipment is likely to be a commercial telephony switch such as System X from Marconi or AXE from Ericsson.

Between the DSL port 264 of DSLAM 26 and second E-side termination 243 of MDF 24, the DSL connection, and hence the test path, passes through TAM 28. For the purposes of this test, TAM 28 does not disrupt the connection which continues to MDF 24, as described above. The TAM 28 contains a number of relays connected in the line and operable to switch so as to break the line into two segments. Each segment so created may then be connected to additional equipment, including test devices, or left open circuit. TAM 28 has a number of further ports 281 to which additional equipment, including test devices, may be connected. The testing is controlled by GTCJ-TE via GTCS and GTCI (all not shown) which connect to test head T/TH 20 and TAM 28.

Figure 2b relates to the same access network as Figure 2a with corresponding features numbered consistently. Figure 2b shows second WW test path from T/TH 20 to TAM 28, that is, in the arrangement of Figure 2b, we see the same equipment as in Figure 2a with a similar test path except that the TAM is switched so as to break the original test path of Figure 2a (as indicated in Figure 2b by the cross in TAM 28). As a result, the test path from test head T/TH 20 terminates at TAM 28.

The test results from the first and second WW test are compared in order to identify inconsistencies between the database contents and the actual configuration of the access network.

In the arrangement of Figures 2a and b, the path of the first test is from T/TH 20 via DSLAM 26 and TAM 28 to the EU (not shown). Before the second line test takes place, TAM 28 is operated by GTCI (not shown) to disconnect the line between DSLAM 26 and E-side connection 243 on MDF 24 such that the path of the line subject to the second test only extends from T/TH 20 to TAM 28. Therefore, the first line test sees the EU termination and a relatively long line while the second test detects (in the absence of any errors) a shorter line due to disconnection before the MDF. As a result of the disconnection, different results are expected from the two tests when the broadband line under test is correctly wired. If the second test returns the same results as the first

test this indicates an incorrect PSTN line connection or a line connected to a wrong TAM port.

The WW test is able to detect reliably if a selected broadband line is correctly or incorrectly wired at the MDF. As shown in Figures 2a and 2b, the WW test involves two line tests which are performed one after the other under the control of GTCS (not shown). These tests include measurements of the characteristics of the line, such as capacitance and resistance in the case of a metallic line. Such characteristics will vary in a measurable way in relation to length of the line, for example, the capacitance should increase with the length of the line. Equivalent tests exist for other media, such as time-domain reflectometry (TDR) which can give a reliable indication of line length for optical paths.

Two versions of the WW test are used according to the embodiments described here. WW(DN) performs a WW test on a line identified by the given DN in an input pair. WW(DSLAM) performs a WW test on a line identified by the given DSLAM port number in an input, pair.

To use the WW test for broadband data verification, it is essential to interpret the test outcomes appropriately. The possible test outcomes are "pass", "fail", "inconclusive" and "test failure". Here, "pass" indicates a correspondence between the physical circuit and the data supplied, whilst "fail" indicates a problem due to the physical circuit failing to match the data supplied. The problem could lie with the circuit or the data. Reasons for returning a result of "inconclusive" include a very short line or an invalid capacitance reading. Reasons for returning a result of "test failure" include no test result due to time-out and DN or DSLAM port number missing from the pairs input to the test.

Table 2, below, shows WW test outcomes on 30 pairs for scenario 2 (DN1, DSLAM porti), (DN 1 , DSLAM portj). In the first row of table 2, test WW(DSLAM porti) - see first column - finds no DN 1 on the DSL database. So, the eight tests provide no information on the pairs (DN1 , DSLAM porti). But, the corresponding eight pairs based on (DSLAM porti) - see second column - have the outcome 'PASS'. The 'PASS' recorded for the (DSLAM porti) pairs prove indirectly that the other eight pairs based on (DSLAM porti) are incorrect.

In the second row, each of two WW(DSLAM porti) tests finds that input DSLAM port value porti is associated in the access circuit with a new DN value DNi different from

the input value DNl This again proves the input pairing (DN1 , DSLAM porti) to be incorrect. Advantageously, VDMT can store such pairs (e.g. DNi, DSLAM porti) newly established as correct by the tests and make use of them to find more incorrect pairs. For example any other pairing of DNi, eg. (DNi, DSLAM portm), can now be flagged as incorrect

In the third row of Table 2, as indicated in column 1 , for three pairs, the DSLAM port number porti is not found in NCAS, and, as indicated in column 2, these three pairs are indirectly proved incorrect by the corresponding three pairs involving DN1 and DSLAM portj having the test outcome 'PASS'.

As indicated in the fourth row of Table 2, in column one, test WW(DSLAM porti) finds a new correct pair (thus proving the first input pair of DN 1 and DSLAM porti to be incorrect). This is corroborated by test WW(DSLAM portj), in the second column.

From the test data displayed in Table 2, the following conclusions can be reached,

1 ) 29 pairs (see first column rows 2 to 5 and column two rows 2 to 4) are on TAM enabled sites and could, therefore, be tested. One pair (see second column, row 5) is on a non-TAM site where the WW test is not available.

2) WW tests establish a definite result, directly or indirectly, for all 29 pairs on TAM enabled sites, i.e. establishing the input pairs to be either correct or incorrect. For the pairs recorded in rows two to five one pair is directly proved to be correct, thus indirectly proving the other pair to be incorrect.

3) Among the 29 tested pairs, the WW tests discover four new correct pairs (i.e. different from the input pairs). These newly discovered, correct pairs can be stored and used to identify more incorrect pairs.

4) Two or three WW tests may be needed in order to provide enough information to categorise the two input pairs.

5) The test data do not show the problem cause for scenario-2 in Table 1 (i.e., the POTS database was not updated with a new DSLAM port following a reassignment of DSLAM ports).

For an input pair (DNi, DSLAM porti), GTCJ-Transfer Engineering performs WW(DNi) and/or WW(DSLAM porti) as necessary. The following lists possible test outcomes by WW.

1 ) 'PASS on DNi, DSLAM porti', i.e., a direct proof of input pair (DNi, DSLAM porti) being correct.

2) 'FAIL on DNi, DSLAM porti', i.e., a direct proof of input pair (DNi, DSLAM porti) being wrong. 3) 'PASS on DNm, DSLAM porti', i.e., a direct proof of (DNm, DSLAM porti) being correct and therefore input pair (DNi, DSLAM porti) being wrong. VDMT can use the newly discovered correct pair (DNm, DSLAM porti) and search for the incorrect pair (DNm, DSLAM portn) in the POTS database and DSL database.

4) 'PASS on DNi, DSLAM portm', i.e., a direct proof of (DNi ₁ DSLAM portm) being correct and (DNi, DSLAM porti) being wrong.

5) 'NOTEST - no DSLAM porti found on NCAS', i.e., an indirect proof of (DNi, DSLAM porti) being incorrect. It is recommended that Transfer Engineering broadband OSS checks NCAS first and finds non-existence of some DSLAM ports. This would allow testing in such cases to be suppressed. 6) 'NOTEST - no DNi (or no DSLAM porti) on the POTS database', i.e., no direct information on this pair. The indirect information should be used, i.e., outcomes from testing the other pair (DNj, DSLAM portj). The appropriate inference rules can be derived from the detailed discussion of Table 2.

7) 'NOTEST - no TAM'. If no TAM present on a circuit, there may exist the possibility of falling back to conventional frame jumper test or manual verification

8) 'NOTEST - time out/line busy/system busy ...', i.e., need a re-test or manual verification

9) 'INCONCLUSIVE - invalid measurements ...', i.e., need manual data verification

The invention proposes a reliable and informative test according to which decision- making is based on detecting a disconnection at the TAM using remote test equipment, rather than using the TAM in the conventional way as a base from which to connect test equipment. An evaluation has shown that WW Test gives reliable 'pass' and 'fail' decisions with only a very small percentage of 'inconclusive' results. It is significantly

more reliable than the conventional OFJ Test which produces a higher percentage of inconclusive results.

The testing is based on two parameters and either both parameters may be given or a single parameter given and the second derived from network databases, as described above.

The exchange equipment whose connection in the access circuit is verified may include a telephone switch, an RCU or other PSTN equipment including a multiplexer, such as a DSLAM. Various items of test equipment may be connected to the ports 281 of TAM 28, including but not limited to A/TH and broadband test equipment.

Table 2. Test Outcomes.

The embodiments described, above, rely on the connection of a TAM in the circuit under test. However, many operators have no TAM installed in their networks so that not all circuits will have a TAM available. The provisioning of a TAM merely for the purposes of the WW test may often be viewed as unjustifiable on grounds of cost. A further variant of the WW test is therefore proposed that does not require a TAM in the circuit.

Rather than breaking the line at the TAM, as in the embodiments described earlier, this variant depends on changing the characteristics of the line at a point local to the DSLAM port. According to a preferred embodiment, a change is introduced into a circuit by reconfiguring a DSLAM connected to that circuit to present a measurable difference in the line characteristics.

At least some commercially available DSLAMs (for example products provided by Fujitsu and Alcatel) include a bypass relay that can be operated according to the present embodiment to change the impedance characteristic of the line. The bypass relay is provided to allow bypassing of the low pass filter (LPF). According to this embodiment, the WW test controller GTC runs the WW test twice via the test head: once with the low-pass filter in the circuit and the other without. According to a preferred embodiment, this WW test may be run by driving the line with test signals at a pair of frequencies and detecting the line characteristics at each frequency of the pair.

A suitable test head is selected for providing the two test signals and monitoring the impedance change.

Figure 3 shows a test arrangement according to a further embodiment of the invention. Components common to earlier figures bear the same reference numerals and will not be described further here. The arrangement shown in Figure 3 is essentially the same as that of Figures 2a and 2b except for the absences of TAM 28. Hence, as shown in Figure 3, the test path from POTS port 262 on DSLAM 26 passes through DSLAM 26 exiting at the user DSL port 264 corresponding to entry POTS port 262 and continues directly back to MDF 24 at second E-side termination 243 without passing through a TAM on the way.

Figure 4 shows DSLAM 26 of Figure 3 in more detail. In Figure 4, some internal details of DSLAM 26 are shown. In particular, the test path internal to DSLAM 26 from user DSL port 264 to POTS port 262 passes through bypass relays 261 and low-pass filter 266. Similarly, the path internal to DSLAM 26 from user DSL port 264 to the separate ATM network (not shown) passes through high-pass filter 268. These filters are present so as to separate the voice-band and DSL signals, with the high-pass filter passing the DSL signals to the ATM network but blocking the voice-band signals. Similarly, the low- pass filter passes the voice-band signals to the PSTN network but blocks the DSL signals. Figure 4 shows relays 261 in the bypass position, with the test path internal to DSLAM 26 from user DSL port 264 to POTS port 262 bypassing LPF 266.

LPF 266 remains in circuit for regular, overnight line tests. The LPF 266 is left in circuit for this test as (i) removal of the LPF 266 would disrupt the customer's broadband service and (ii) the capacitance added by the LPF 266 to the line is not seen as adversely affecting the results of this test. The overnight test, where the LPF 266 remains in circuit, is referred to as a non-intrusive test. Where a fault is reported in a customer's line, a different test (referred to as an intrusive test) is required that uses the measured value of line capacitance in order to estimate the location of the fault along the line. For this test, the presence of the LPF 266 in the line under test can significantly affect the results, leading to an inaccurate estimation of the location of a fault. For this reason that the by-pass relay is switched to remove the LPF 266 from the circuit for this test.

According to the present embodiment, the bypass relay is used to provide additional information about the line circuit, not available from the conventional tests, described above. In the arrangement of Figure 4, according to the invention, the first WW test is performed with LPF 266 in circuit within the DSLAM 26 between DSL port 264 and corresponding POTS port 262 such that the electrical characteristics of the line subject to the first test are influenced by the presence of LPF 266. Before the second WW line test takes place, DSLAM 26 is instructed by GTCS (not shown) to switch LPF 266 so as to short circuit or bypass it or disconnect it from the line within DSLAM 26, such that the electrical characteristics of the line subject to the second test are not influenced by the presence of the LPF 266.

The test head is configured to measure the line impedance, including a complex impedance element, and to detect changes caused by switching of the filter. As a result of switching LPF 266, different results are expected from the two tests when the broadband line under test is correctly configured. If a line is wired correctly from a POTS port (identified by the input DN) through a correct DSLAM port (identified by

CBUK), the first test detects LPF 266 while the second sees no LPF. If the second test returns the same results as the first test, this indicates an incorrect PSTN line connection or a line connected to a wrong DSLAM port. The test head is configured to return to the GTC results of the test from which it can be determined whether the LPF 266 has been detected or not.

The interface between the WW test controller GTC and test head is largely similar to that used for the copper line test except for provision of an extra message field indicating if a LPF is detected or not. LPF WW test may be run in coordination with a copper line test followed by DSLAM detection. The LPF WW test does not require the presence of a TAM in the line under test but may be applied to TAM lines in addition to, or in place of, the TAM-based WW tests described above.

The present invention exploits bypass relays 261 for the purpose of testing for the presence of a DSLAM in a PSTN line. Low-pass filter (LPF 266) is normally attached to the line by the DSLAM in order to filter out voice-band signals (i.e. PSTN telephony signals) received from the line. The filter may also function as a path to insert voice- band signals into the line. Controller GTCS interfaces with the DSLAM EM (Element Manager of the DSLAM) to issue a command to directly to the bypass relay for the low- pass filter identified by a specific DSLAM port number.

As will be understood by those skilled in the art, the invention may be implemented in software, any or all of which may be contained on various transmission and/or storage mediums such as a floppy disc, CD-ROM, or magnetic tape so that the program can be loaded onto one or more general purpose computers or could be downloaded over a computer network using a suitable transmission medium. The computer program product used to implement the invention may be embodied on any suitable carrier readable by a suitable computer input device, such as CD- ROM, optically readable marks, magnetic media, punched card or tape, or on an electromagnetic or optical signal.

Those skilled in the art will appreciate that the above embodiments of the invention are greatly simplified. Those skilled in the art will moreover recognise that several equivalents to the features described in each embodiment exist, and that it is possible to incorporate features of one embodiment into other embodiments. Where known equivalents exist to the functional elements of the embodiments, these are considered to be implicitly disclosed herein, unless specifically disclaimed. Accordingly, the spirit and scope of the invention is not to be confined to the specific elements recited in the description but instead is to be determined by the scope of the claims, when construed in the context of the description, bearing in mind the common general knowledge of those skilled in the art.

The content of the attached abstract is incorporated herein, as follows. A method and tester for verification of a set of access circuit parameters indicating an access circuit allocated to a user; and an equipment port on an item of equipment. The method includes the steps of (a) testing the access circuit associated with the user; (b): applying a condition to an access circuit at a point between a user and the equipment port and (c) testing again the access circuit tested in step (a) so as to determine if the equipment port is connected in the access circuit associated with the user. Each circuit is arranged to carry an instance of a first and an instance of a second communications service. Each service comprises a plurality of instances with each instance being associated with one of the circuits. The tester comprises means for performing a first and a second test on the circuit associated with a first instance of the first service; means for applying a condition to the circuit associated with a first instance of the second service during the second test but not during the first test and means for analysing the results of the first and second tests to detect in the circuit associated with the first instance of the first service the condition applied during the second test.

GLOSSARY

21CN 21 ^st Century Network

A/TH Acterna Test Head (e.g. QT-200 DSL Test Solution, see http://www.idsu.com/test and measurement/technical resources/prod uct documents/datasheet/qt200 ds cab tm ae 0106 2nd.pdf )

ADSLEM ADSL Element Manager, such as the Fujitsu FENS-AN (see: http://www.fujitsu.com/uk/services/telecom/products/network/ ) and the Alcatel AWS (see: http://www.alcatel.com/products/productsummary.jsp?relativeP ath=/co m/en/appxml/opgproduct/alcatel5523adslworkstationtcm22811797 163 δ.jhtml )

BB broadband

CBUK Broadband network identity

CSS Customer Services System = the POTS database

DN Directory Number [BT]

DSL Digital Subscriber Line

DSLAM Digital Subscriber Line Access Multiplexer

ETAM Exchange Test Access Matrix

EU End User

ETAM Evolutionary Test Access Matrix

FTAM Frame Test Access Matrix

GTC Generic Test Controller

MDF Main Distribution Frame

NISM Network Inventory and Spares Management

NCAS Network Capacity Assignment System

NT Network Terminating Equipment

OSS Operational Support System

POTS Plain Old Telephony Service

PSTN Public Switched Telephone Network

RCU Remote Concentrator Unit (System X)

SIDB Service Inventory Database

TAM Test Access Matrix (e.g. UTEL XDSL Modular Test Access Switch, see: http://www.utelsw.f2s.com/testaccessswitches.shtml )

TAN Test Access Number T/TH Teradyne Test Head from Teradyne Inc. Boston, USA ( see: http://www.teradyne.com/prods/btd/btd products/ldu100/ldu1 OO.pdf )

WW Test TAM-based xDSL jumper test