AU696350B2 - Optical fibre management system - Google Patents

Description

~11111*~- OPI DATE 27/03/95 APPLN. ID 75446/94 lilllllli li lll111 AOJP DATE 18/05/95 PCT NUMBER PCT/GB94/01960 Ii l 1 11 I iii AU9475446 JT) (51) International Patent Classification 6: (11) International Publication Number: WO 95/07486 G02B 6/44 SAl f (43) International Publication Date:

I

16 March 1995 (16.03.95) (21) International Application Number: (22) Internatinnal Filing Date: 8 S< PCT/GB94/01960 er 1994 (08.09.94) eptemb (81) Designated States: AU, CA, CN, JP, KR, NZ, European patent (AT, BE, CH, DE, DK, ES, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE).

Published With international search report.

Priority Data: 93307145.8 10 September 1993 (10.09.93) EP (34) Countries for which the regional or international application was filed: AT et al.

94303091.6 28 April 1994 (28.04.94) EP (34) Countries for which the regional or international application was filed: AT et al.

(71) Applicant: BRITISH TELECOMMUNICATIONS PUBLIC LIMITED COMPANY [GB/GB]; 81 Newgate Street, London EC1A 7AJ (GB).

(72) Inventors: FROST, Peter, Lewis, John; 14 Copswood Close, Kesgrave, Ipswich, Suffolk IP5 7QF KERRY, John; 4 The Grove, Martlesham Heath, Ipswich, Suffolk IP5 7UZ PEACOCK, John; 14 Spring Park, Otley, Suffolk IP6 9NX (GB).

(74) Agent: PRATT, David, Martin; BT Group Legal Services, Intellectual Property Dept., 13th floor, 151 Gower Street, London WC1E 6BA (GB).

(54) Title: OPTICAL FIBRE MANAGEMENT SYSTEM So (57) Abstract An optical fibre management system for managing a plurality of optical fibres. The system is such that th: fibre(s) of each single circuit is/are routed separately from the fibres of other circuits, whereby optical signals carried by any given single circuit are not degraded by handling operations carried out on the fibres of other circuits.

-I I a dLI IF D~I~ILIIBPIIIIBB~PB119~ 13llmn~4lso~-ra~,,, WO 95/07486 WO 95/07486PC7I011941) t96() OPTICAL FIBRE MANAGEMENT SYSTEM This invention relates to an optical fibre management system, and in particular to an optical fibre splitter array sub-assembly for incorporation in the node of an optical fibre telecommunications network.

In the United Kingdom, the telecommunications network includes a trunk network which is substantially completely constituted by optical fibre, and a local access network which is substantially completely constituted by copper pairs. Flexibility in the copper access network is provided at two points en route to the customer; firstly, at streetside cabinets serving up to 600 lines; and secondly, at distribution points serving around 10-15 lines. In total, the network has about 250,000km of underground ducts, 83,000 cabinets, 3. 1 million distribution points and 3.7 million manholes and joint boxes. Eventually, it is expected that the entire network, including the access network, will be constituted by fibre.

The ultimate goal is a fixed, resilient, transparent relecommunications infrastructure for the optical access network, with capacity for all foreseeable service requirements. One way of achieving this would be to create a fully-managed fibre network in the form of a thin, widespread overlay for the whole access topography as this would exploit the existing valuable access network infrastructure. Such a network could be equipped as needs arise, and thereby could result in capital expenditure savings, since the major part of the investment will be the provision of terminal equipment on a 'just in time' basis.

It should also enable the rapid provision of extra lines to new or existing customers, and flexible provision or reconfiguration of telephony services.

In order to be completely future proof, the network should be single mode optical fibre, with no bandwidth limiting active electronics within the infrastructure.

Consequently, only passive optical networks (PONs) which can

I

~R~~i WO 95/07486 PCTIGB)94/0 1960 2 offer this total transparency and complete freedom for upgrade, should be considered.

The most common passive optical network is the simplex single star, with point-to-point fibre for each transmit and receive path, from the exchange head end (HE) to the customer network terminating equipment (NTE). This network design has been used throughout the world and meets all the access criteria. It involves high fibre count cables, and unique electro-optic provision at HE and NTE for each customer. The resulting inherent cost can only be justified for large business users, who generally also require the security of diverse routing, which increases the cost still further.

The advent of optical splitters and wavelengthflattened devices has enabled the concept of the PON to be taken one step further. These passive components allow the power transmitted from a single transmitter to be distributed amongst several customers, thereby reducing and sharing the capital investment. In 1987, BT demonstrated splitter technology in a system for telephony on a passive optical network (TPON), with a 128 way split and using time division multiplex (TDM) running at 20Mb/s. This combination enabled basic rate integrated service digital network (ISDN) to be provided to all customers. In practice, the competitive cost constraint of the existing copper network precludes domestic customers from having just telephony over fibre, due to the high capital cost of equipment. This may change in the future. In the meantime, telephony for small business users (for example those having more than 5 lines) can probably break this barrier.

The wider range of services and higher capacity required by business customers makes a 32-way split more attractive for a 20 Mb/s system and this has been demonstrated by BT's local loop optical field trial (LLOFT) at Bishop's Stortford.

In summary, the use of splitter based PON architecture will reduce the cost of fibre deployment in the access network. When compared with point-to-point fibre, savings ~L I I I ww R R BIO W -n l WO 95107486 ICT/G 94/O 1 96(0 will result from: reducing the number of fibres at the exchange and in the network; (ii) reducing the amount of terminal equipment at the exchange; (iii) sharing the cost of equipment amongst a number of customers; (iv) providing a thin, widespread, low cost, fibre infrastructure; and providing a high degree of flexibility, and allowing 'just in-time' equipment and service provision.

Additionally, PON architecture can be tailored to suit the existing infrastructure resources (duct and other civil works).

Total network transparency will retain the option for future services to be provided on different wavelengths to the telecommunications, which for TPON is in the 1300nm window. By transmitting at other wavelengths, other services, such as broadband access for cable television and high definition television, or business services, such as high bit rate data, video telephony or video conferencing, can be provided. The huge bandwidth potential of fibre promises virtually unlimited capacity for the transparent network. Eventually, it will be possible to transmit hundreds of wavelengths simultaneously, as the development of technology in optical components, such as narrow band lasers, wavelength division multiplexers (WDMs), optical filters, fibre amplifiers and tunable devices, moves forward.

For this potential to remain available, and for the access network to be used to provide many and varied services, then it must be designed and engineered to provide very high levels of security and resilience. Even for simple POTS, advance warning and live maintenance are essential to limit disruption.

Resilience implies separacy of routing, and exploiting the existing infrastructure of underground ducts and other civil works is a prime requirement of the design philosophy.

-I I I IsAsPRB1%IEllls4(~0WIBR~RIBIRI~~~" WO 95/107,186 ,1;/01960 4 Analysis of this resource has indicated that separacy, from creating primary ring topographies, could be achieved by linking the spine cables which currently feed many primary connection points (PCPs) in the existing star type network.

In order to create rings from the existing star configurations, some localities will have existing ducts that will allow the link cables to be installed. In BT' s suburban networks, analysis has shown that on average 60% of PCPs can be served on rings using existing ducts; and, by adding new ducts links of 200m or less, a further 30% can be covered.

In some cases, there will be natural or man made boundaries where physical rings cannot be provided, and in these cases duplicate fibres in the same duct route, i.e. across rivers or over railway bridges, may be the only choice.

The architecture adopted for the PON topography will be influenced by transmission techniques, and the availability of suitable splitter components. Transmission options are simplex (two fibre paths), duplex, half duplex or diplex (single fibre path).

Simplex working increases the complexity of the Infrastructure due to the two fibres per circuit required.

However, it benefits from the lowest optical insertion loss, due to the absence of duplexing couplers; and the lowest return loss, since such systems are insensitive to reflections of less than 25 dBm with separate transmit and receive paths. Duplex and half duplex working each have an -nsertion loss penalty of 7dB from the duplexing couplers, and diplex working replaces these with WDMs, with a reduced penalty of 2dB.

In view of the long term aim to provide a total fibre infrastructure, and the present early state of passive technology components, it is considered beneficial to opt for simplex working and a relatively low level of split for PON networks.

In an optical fibre communications system, transient changes in optical attenuation can cause transmission errors.

These changes are caused by transient bend loss at various -iM *IIR*IBII~L*~ RI~IPPR~~~ WO 95/07,16 I"/G')TO941960O 5 points along the fibres of the system, and the extent to which traffic along a given fibre is disturbed is dependent on such physical variables as the total loss incurred and the duration of the transient. Transient losses occur mainly because of fibre handling and maintenance procedures, particularly in the regions of fibre splices. Thus, when multi-fibre splice trays are opened, and/or fibre is handled, attenuations of up to 10 dB can be observed. For example, a typical splice tray used in optical communications systems contains 24 splices, and handling any one of the splices for maintenance purposes causes transient losses in adjacent fibres. This problem is illustrated in Figure 15 of the accompanying drawings, which plots the probability of the occurrence of error against the system margin, at both 1550nm and 1300nm for a procedure which includes opening a typical 24 fibre splice tray, and running a finger along the splices.

Error loss measurements are made of the fibres at splice position 14, as this splice position is almost in the centre, and so is more susceptible to transient losses than other splice positions. As shown in Figure 15, at the optimum operating position of the receiver making the error measurement (that is to say at a system margin of 0 dB) at both 1550nm and 1300nm there is a large percentage error occurrence as a result of the transient losses caused by the fibre handling. As the system margin is increased, the percentage error occurrence falls at both 1550nm and 1300nm, but there is still a significant percentage error occurrence at 1550nm even as the system margin approaches the dynamic range (typically 15 dB) of the receiver. The normal operating position of the receiver is the optical power nominally detected at the receiver to achieve a bit error rate (BER) of 10'9 or better. The results at 1550nm are far worse than those at 1300nm, due to increased bend sensitivity at 1550nm and hence larger transients. This is potentially worrying if splice trays are to be installed in a system operating at 1300nm with later provision for operation at 1550nm. This is because there may be a point at which a i I -6system may operate with no handling errors at 1300nm, but will show a serious handling error performance at 1550nm due to the increased bend loss sensitivity of fibres at 1550nm. This would lead to a need for an increased system margin at 1550nm to compensate for the greater losses at 1550nm. This is undesirable because it results in lower power budgets due to a higher required incident optical power.

The aim of the invention is to provide a fibre management system which does not suffer from the type of transient losses referred to above. This is accomplished by what may be termed "single circuit management". A single circuit may be constituted by a single fibre which connects a transmitter/receiver pair at a first location and a transmitter/receiver pair at a second location. Similarly, a single circuit may be constituted by a first fibre connecting a transmitter at a first location to a receiver at a second location, and a second fibre connecting a transmitter at the second location to a receiver at the first location. Again, a single circuit may be constituted by a plurality of optical fibres interconnecting a transmitter at a first location and a receiver at a second location, and a plurality of fibres connecting a transmitter at the second location and a receiver at the first location.

The optical fibre communications system of the invention provides single circuit management and so ensures that optical signals carried by any given single circuit are not degraded by installation/maintenance operations carried out on other single circuits of the system. This is accomplished by ensuring that each single circuit is housed and routed as a separate entity at a point in the network where reentry for installation/maintenance purposes is possible.

Thus, according to a first aspect, the present invention provides a method of providing single circuit management within a network node of a passive optical network, the method including the steps of: routing input optical fibre(s) from a cable element to an input splice tray; splicing said input optical fibre(s) to optical splitter input fibre(s) in said input splice tray, said optical splitter input fibre(s) being optically coupled to an optical splitter, said optical splitter being optically coupled to a plurality of optical splitter output fibres and each of the optical splitter output fibres being spliced to a respective output optical fibre within a respective output splice tray; and routing each of the output optical fibres from their respective output splice tray to a respective output optical fibre cable or a respective output optical fibre cable element such that the input optical fibre(s), the splitter input fibre(s), the splitter output fibres and the output optical fibres are not bent to a radius less than the minimum bend radius of live optical fibre, wherein the optical fibre(s) forming each single circuit is separately routed through a dedicated input splice tray and a dedicated output splice tray such that the optical fibre(s) forming each single circuit can be accessed within said dedicated input splice tray or said dedicated output splice tray without causing substantial disturbance to the optical fibre(s) forming another single circuit. Preferably the input optical fibre is routed from the cable element to the input splice tray via a bend limiting tube, and each output optical fibre is routed from the output splice tray to the respective output optical fibre cable or the respective output optical fibre cable element via a respective bend limiting tube.

According to a second aspect of the invention, there is provided a method of providing single circuit management within a network node of an optical network, the method including the steps of: routing an input optical fibre from a cable element to a splice tray; splicing said input optical fibre to an output optical fibre within said splice tray; and routing said output optical fibre from said splice tray to an output optical fibre cable or an output optical fibre cable element such that the input optical fibre and the output optical fibre are not bent to a radius less than the minimum bend radius of live optical fibre, wherein the optical fibre(s) forming each single circuit is separately routed through the splice tray such that the optical fibre(s) forming each single circuit can be accessed within said splice tray without causing substantial disturbance to the optical fibre(s) forming another single circuit.

Preferably, the input optical fibre is routed from the cable element to the splice tray via a bend limiting tube, and the output optical fibre is routed from the splice tray to output optical fibre cable or the output optical fibre cable element via a bend limiting tube.

According to a third aspect of the invention, there is provided an optical fibre management apparatus for use in a node of an optical fibre network comprising a cable element storage area, one or more input splice trays, one or more optical splitters and one or more output splice trays, the optical fibre management apparatus being configured to: guide an input optical fibre from a first cable element stored in the cable element storage area to one of the input splice trays; guide an optical splitter input fibre from said input splice tray to one of the optical splitters; guide an optical splitter output fibre from said optical splitter to one of the srta~B ~rarr~R~asl 8 l a~-~uol- 7a output splice trays; and guide an output optical fibre from said output splice tray to a second cable element stored in the cable element storage area such that the input optical fibre, the optical splitter input fibre, the optical splitter output fibre, the output optical fibre, the first cable element and the second cable element are managed such that they are not bent to a radius less than the minimum bend radius of live optical fibre. Preferably, the apparatus additionally comprises at least one length of bend-limiting tubing, the or each length of bend-limiting tubing guiding at least one of the input optical fibre, the optical splitter input fibre, the optical splitter output fibre and the output optical fibre such that they are not bent to a radius less than the minimum bend radius of live optical fibre. The one or more input splice trays and the one or more output splice trays may be arranged in a vertical stack.

Unless the context clearly requires otherwise, throughout the description and the claims, the words 'comprise', 'comprising', and the like are to be construed in an inclusive as opposed to an exclusive or exhaustive sense; that is to say, in the sense of "including, but not limited to".

The invention will now be described in greater detail, by way of example, with reference to the accompanying

I

wVo 4A07,W10 190(t drawings, in which: Figure 1 is a perspective view, for one side, of an optical fibre telecommunications network node incorporating three splitter array sub-assemblies each of which is constructed in accordance with the invention; Figure 2 is a perspective view, form the opposite side, of the node of Figure 1; Figure 3 is a perspective view showing the node of Figures 1 and 2 mounted in a footway box in a storage position; Figure 4 is a perspective view similar to that of Figure 3, but showing the node 2 mounted in the footway box in its working position; Figure 5 is an exploded perspective view of one of the splitter array sub-assemblies of the node of Figures 1 and 2; Figure 6 is a perspective similar to that of Figure but showing parts of the sub-assembly, then parts being in their operative positions; Figure 7 is a perspective view of one of the splice trays of the splitter array sub-assembly of Figures 5 and 6; Figure 8 is a plan view showing the fibre entry/exit portion of the splice tray of Figure 7.

Figure 9 is a perspective view of one of the bendlimiting tube manifolds of the splitter array sub-assembly of Figures 5 and 6; Figure 10 is a perspective view of one of the coupler array mats of the splitter array sub-assembly of Figures and 6; Figure 11 is a perspective view of the coupler array back cover of the splitter array sub-assembly; Figure 12 is a plan view of the break-out tray which forms part of the node of Figures 1 and 2; Figure 13 is an enlarged perspective view of part of the break-out tray; Figure 14 is a perspective view of another form of splice tray which can be incorporated into a splitter array sub-assembly; and I II I Ll ~.LI

CIW~--

WO90,1'10W7 i( TI(; /lif,Mt I 9l Figure 15 is a graph plotting error probability against system margin for a known type of splice tray.

Referring to the drawings, Figures 1 and 2 show a node N forming part of a ring topography PON, The node N includes Sa stack of three splitter array sub-assemblies S, and S 3 and a break-out tray T. A 96 fibre cable C, which forms a ring (loop) centred on a local exchange (not shown), enters the break-out tray T via a cable entry portion 2 (see Figure 12) after passing through a node base 1. The cable (then passes at least twice round a generally oval perimeter track 3 of the tray T, and leaves the tray via the portion 2. The 96 fibres are housed in twelve flexible tubes (not shown) made of plastics material, each of the tubes containing eight primary-coated fibres. As is described in detail below with reference to Figure 12, the tray T includes a break-out region B in which individual fibre end portions, formed by cutting into one of the tubes, are led to the splitter array sub-assemblies S, S 2 and S 3 In this connection, it should be noted that the tray T stores a sufficient length of the cable C so that, after cutting one of the tubes in the middle of this stored length, and stripping back that tube to expose _ts optical fibres, each of the originally continuous fibres forms two fibre end portions whose length is sufficient to be led to the splitter array sub-assemblies S, and S 3 and to leave spare fibre which can be stored for future use.

Figures 3 and 4 show the mounting of the node N in a footway box F, a dome-shaped cover D being fixed to the node base 1 prior to mounting.

One of the splitter array sub-assemblies, S, is shown in detail in Figures 5 and 6. The other two sub-assemblies S. and S, are the same as the sub-assembly S. The subassembly S. includes a stack of ten splice trays 4, each of which _s 8mm thick. The trays 4 are supported (in a manner described below) by a stainless steel chassis 5 constituted by a top plate 5a, a base plate 5b and a back plate 5c. Each of the splice trays 4 is a single circuit splice tray, that is to say, in use, it has two incoming optical fibres (one L~JP~er-- M WVO 90/07110 P4 /(94I t il Uci :r L 3 lIi 4 t Itg l a ucu.oa vlo I V I and t wo outqo tin )ptical fibros (one each for transmittinq and rec ving).

The three plates Sa, 5b and 5c are welded together, a'd the top plate 5a of the sub-assembly S, can be fixed to the base plate 5b of the adjacent sub-assembly S, (not shown in Figures and 6) by means of mounting bolts (not shown). Similar mounting bolts can be used to fix the plate 5a of the subassembly S. and the plate 5b of the sub-assembly S 3 to support means (not shown) in the node N.

'z.e chassis 5 also supports an input splitter array mat 6, an output splitter array mat 7, and a splitter array back cover 8. In this connection, the input mat 6 carries (as is described below with reference to Figure 10) fibres which carry telecommunications signals from the exchange to customers. These fibres are termed transmit fibres.

Similarly, the output mat 7 carries fibres which carry telecommunications signals from customers to the exchange.

These fibres are termed receive fibres. The mats 6 and 7 are made of a flexible polymer, for example an elastomeric polymer such as injection mouldable zantoprene, or polyurethane. The back cover 8 is made of flexible polypropylene (which is also injection mouldable). This inherent flexibility ensures that, in use, the mats 6 and 7 are held firmly against the chassis back plate 5c by the back cover 8.

As shown in Figure 7, each splice tray 4 has a main body portion 9 and a fibre entry portion 10 which also constitutes a clip-on test area. Fibre access to the main body portion 9 from the fibre entry portion 10 is via a channel 11. The main body portion 9 is of oval configuration, having an oval base 9a and an upstanding peripheral wall 9b. A hollow mandrel 12 is provided on the base 9a adjacent to the entry channel 11. The mandrel 12 is of rounded square cross-section, is sized to ensure minimum bend requirements for live fibre passing around it, and has a fibre inlet aperture 12a through which dark fibre can pass for internal storage. A channel 13 is defined between the I~ I -I WO 9 107,186 PCI(G194/019611 manarel 12 and the peripheral wall 9b, the channel 13 leading to a further channel 14 which leads around the inside of the wall to a splice holder region 15. In use, this region houses a splice holder (not shown) for splicing two incoming fibres to two outgoing fibres. A direction reversing channel 16 leads from the channel 14 adjacent to the region 15 back to that portion of the channel 14 which adjoins the cnannel 13 adjacent to the mandrel 12.

The fibre entry portion 10 of each splice tray 4 includes three fibre entry/exit ports 17a, 17b and 17c (see Figure Diverging channels 18a and 18b are provided to lead fibre between the port 17a and the channel 11 via respective apertures 19a and !9b. These apertures 19a and 19b constitute what are known as "clip-on apertures", and provide easy access to the associated fibres in order to measure the light passing there along, and hence to determine the quality of the splices. These clip-on apertures, and associated light measurement apparatus, are described in the specification of our International patent application WO 93/00600.

Similar diverging channels 20a and 20b are provided to lead fibre between the port 17c and the channel 11 via respective clip-on apertures 21a and 21b. A single channe' 22 i. provided for leading fibre between the port 17b and t., channel 11. The channel 22 Is not provided with a clip-on aperture.

Each splice tray 4 is also provided with a number of fibre retention tabs 23 for holding fibre in the various channels 11, 13, 14, 16, 18a, 18b, 20a, 20b and 22. One of these tabs (indicated by the reference numeral 23a) is generally V-shaped, and extends from the curved end of the peripheral wall 9b remote from the mandrel 12 about halfway across, and above, that portion of the base 9a between that wall portion and the mandrel.

Each tray 4 is pivotally mounted on the splitter array back cover 8 by means of a leash 24 and a retaining ring which are moulded integrally with the rest of the tray. The as3 r~r r;n RIIIQLI~B~-anbblllll1~4lr~ ~l- WO 95/0746H 1M(11t1GIR911/O1 i) 12 leash 24 of each tray 4 has two arms a and 24b joined together by a hinge 24c. Its retaining ring 25 is a friction fit within a groove 26 formed in the back cover 8 (see Figure 11). In use, a rod (not shown) passes through all the retaining rings 25 and through apertures (not shown) in the top and base plates 5a and 5b. In this way all the splice trays 4 are retained by their back plates 5c, but each can be pivoted out away from the other trays in the stack to provide access to its clip-on apertures 19a, 19b, 21a and 21b. In this position, the arms 24a and 24b take up a generally straightline configuration (as opposed to the V-shaped configuration they have when the tray is in the stack). As the retaining ring 25 of a pivoted-out tray 4 is held in position by the retaining rod, the pivotal movement of the tray is limited by the leash 24 as its two arms 24a and 24b straighten out. In the fully pivoted-out position (the first operating position), the fibre entry portion 10 of a tray 4 is exposed.

Each of the splitter array sub-assemblies S 2 and S 3 is associated with two fibres (four fibre end portions) of the eight fibres in the cut tube of the cable. The remaining two fibres (four fibre end portions) from the cut tube are stored in the break-out tray T as is described below with reference to Figure 13. As the cable C is in the form of a ring, telecommunications signals can travel to/from the exchange in either direction round the ring. For convenience, one of the directions is termed the main direction, and the other the standby direction. In practice, only main fibres will be used for normal signalling, the standby fibres only being used in the eventuality of main fibre failure.

The two main fibre end portions associated with say the splitter array sub-assembly S, pass from the break-out tray T to the lowest splice tray 4 of that assembly, the fibre end portions being supported in, and protected by, a bend limiting tube 27a (see Figure This bend limiting tube 27a is a proprietary item made of polypropylene ringed rM WO 9M/07IHb (C 11 tubing which, though flexLble, cannot easily be bent beyond minimum bend radius requirements for live fibre. The bend limiting tube 27a terminates in the port 17a of the lowest splice tray 4, and its two fibre end portions are led into the main body portion 9 via the channels 18a and 18b, the clip-on apertures 19a and 19b, and the channel 11. These fibre end portions are then spliced to the ends of a pair of fibres which (as is described below) are associated with the mats 6 and 7. The two splices are then positioned in a splice holder which is then mounted in the region 15. The four fibres leading to the splices are then stored in the main body portion 9 of the tray 4 with two of the fibres (for example those from the break-out tray T) being led away from the splices in the channel 14, and the other two fibres being led away from the splices via the channel 13 and the reversing channel 16. A length of each of the fibres is stored in the main body portion 9 of the tray 4 by passing these fibres one or more times round the mandrel 12 and under the V-shaped tab 23a. The fibres' natural resilience will ensure that the loops of fibre expand outwardly into a configuration of varying diameter turns. The provision of stored fibre permits a minimum of ten re-eplices of each of the splices to be carried out during the lifetime of the assembly.

The two fibres which are associated with the mats 6 and 7 leave the main body portion of the tray 4 via the channel 11. They are then led to the port 17c of the entry portion 10 via the clip-on windows 21a and 21b and the channels 20a and 20b. These fibres are then led to the mats 6 and 7 within a bend limiting tube 27c (see Figure One of these main input fibres terminates on the input mat 6, where (as is described below with reference to Figure 10) it is joined by splitter means to eight output fibres.

Similarly, the other of these main input fibres terminates on the output mat 7, where it is joined by splitter means to eight output fibres.

The two main fibre end portions and the associated i I I pair of fibres to which they are spliced constitute a single circuit, this circuit starting in the break-out region B of the break-out tray T (a first fibre separation point) and finishing at the input to the splitter mats 6 and 7 (a second fibre separation point. Throughout the length of this single circuit, its fibres are routed separately from the fibres of other circuits, so that single circuit management results. Thus, in the break-out region B of the break-out tray T the two main fibre end portions are separated (as described in greater detail below with respect to Figure 13) from the other cut fibre end portions. These two fibre end portions are then fed to the splitter array sub-assembly S within the bend limiting tube 27a, after which they are fed into the lowest splice tray 4 of that assembly. The two fibres spliced to these main fibre end portions are then fed to the splitter mats 6 and 7 within the bend limiting tube 27c.- Clearly, the fibres carried by this splice tray 4 form part of the single circuit, Consequently, the entire single circuit between the break-out point and the splitting point is separated from all other circuits in the region where re-entry for installation/maintenance purposes is to be expected. This ensures that optical signals carried by this circuit are not degraded by installation/maintenance operations carried out on other circuits of the system.

The two standby fibre end portions associated with this splitter array subassembly S1 pass from the break out tray T to the second lowest splice tray 4 of that assembly. Here, these two fibre end portions are spliced to two fibres which are led back to the mats 6 and 7 and so are termed standby input fibres, and each of the standby input fibres is joined by splitter means to the same eight output fibres as the corresponding main input fibre. The fibre arrangement on this second lowest splice tray 4 is the same as that for the lowest splice tray. Similarly, fibres enter and leave this splice tray 4 in bend limiting tubes 27a and 27c.

Thus, the two standby fibre end portions and the associated pair of fibres to which they are spliced also constitute a single circuit, this circuit starting in the break-out region B of the break-out tray T a first fibre separation point) and finishing at the input splitter mats 6 and 7 a second fibre separation point). As with the single circuit associated with the main fibre end portions, this single circuit has its fibres routed separately from other circuits throughout its length, so that single circuit management results.

'AMENDED SHEET 6i o I, _I The remaining eight splice trays 4 in the sub-assembly S, of Figures 5 and 6 are customer splice trays, As the fibre arrangement in each of these customer splice trays 4 is the same, this will be described in detail for only one of these trays. Thus, one of the output fibres from each of the mats 6 and 7 (that is to say a transmit fibre and a receive fibre) is led to the port 17c of a given customer splice tray 4 inside a bend limiting tube 27c. These two fibres are led into the main body portion 9 of the tray 4 via the channels 20a and 20b, the clip-on windows 21a and 21b, and the channel 11. In use, these fibres are spliced to two fibres of a four-fibre blown fibre unit associated with a given customer. Such a unit has four fibres in a single tube, the tube being fed between the customer and the node N by the well known fibre blowing technique (see EP 108590). The customer's blown fibre unit is led to the port 17a of the splice tray 4 within a bend limiting tube 27a. The blown fibre coatings are stripped away from the four fibres "downstream" of the port 17a.

Two of the fibres within the unit (the two fibres which are to be spliced to the transmit and receive fibres from the mats 6 and 7, and so are termed live fibres) are fed to the main body portion 9 of the splice tray 4 via the channels 18a and 18b, the clip-on apertures 19a and 19b, and the channel 11. The two other fibres (which are spare fibres not for immediate use) are fed to the main body portion 9 of the splice tray 4 via the channels 22 and 11. All four fibres then pass round the mandrel 12 within the channel 13, and then back to the mandrel after passing along the channels 14 and 16. The two spare (dark) customer fibres pass through the aperture 12a and are stored inside the mandrel 12. The two live fibres pass round the mandrel 12, and are then spliced to the transmit and receive fibres from the mats 6 and 7, the splices are stored in a splice holder, and the splice holder is positioned in the region 15. As with the two lowest splice trays 4, each of the spliced fibres has a length to be stored (enabling up to ten re-splices to be made during the lifetime of the assembly), these fibre lengths likewise being stored by looping them each one or more times round the mandrel 12 and under the Vshaped tab 23a.

The two fibres leading to each of the customer splice trays 4 and the two fibres spliced thereto from the associated four-fibre blown fibre unit constitute a single circuit, this circuit starting at the output of the splitter mats 6 and 7 (a first AMENDED SHEET LI III s u fibre separation point), and terminating at the transmitter/receiver pair of the customer. Thoughout the length of this single circuit, its fibres are routed separately from the fibres of other circuits, so that single circuit management results. Thus, from the exit of the splitter mats 6 and 7, the output fibres from the splitter mats are fed to the splitter array sub-assembly S 1 within the associated bend limiting tube 27c. Within the associated splice tray 4, these fibres are spliced to two fibres of the customer's four-fibre blown fibre unit. This tray 4 thus houses only fibres forming part of a single circuit, and the tube of the blown fibre unit separates the fibres of that circuit from fibres of other circuits of the system all the way to the customer's premises. Consequently, the entire single circuit between the output of the splitter mats 6 and 7 and the transmitter/receiver pair of the customer is separated from all other circuits. In particular, in the region where re-entry for installation/maintenance purposes is to be expected (that is to say in the region of the node single circuit management is ensured, so that optical signals carried by any given single circuit are not degraded by installation/mairtenance operations carried out on other circuits.

In order to access the splices within a given splice AMENDED SHEET II wO 'WSn74i( WIWWG)KI/0W900 17 tray 4, it is necessary to remove the rod holding the retaining rings 25 in position, prior to the pulling that tray out of the stack sufficiently far to gain access to the splices. In this position (the second operating postion), the tray 4 is maintained in position by its bend limiting tubes.

The two spare customer fibres stored within the mandrel 12 of a given splice tray 4 can be used to replace the two live fibres of that customer in the event of these fibres failing. More importantly, however, they can be used to provide that customer with additional lines or service.

(In this connection, it should be noted that each fibre pair can provide up to 32 lines using customer premises equipment (CPE) electronics such as an optical network unit (ONU) matched to an optical line terminal (OLT) at the exchange.

Each pair of fibres can also provide a Megastream service.) In this case, the two spare fibres are removed from their storage position .within the mandrel 12, and are led to the fibre entry portion 10 of the tray 4 via the channels 13 and 11. They then leave the tray 4, via the apertureless channel 22 and the port 17b, and enter a bend limiting tube 27b (see Figure This tube 27b is routed via the back cover plate 8 to another splice tray 4 usually a splice tray of another of the sub-assemblies S 2 or S 3 of the node N. The tube 27b terminates at the port 17a of this tray 4, and the two fibres are led into the main body portion 9 via the channels 18a and 18b, the apertures 19a and 19b, and the channel 11. Here they are spliced to two "exchange" fibres, and all spare lengths of fibre are stored in the same manner as that described above for the other splice trays. In this connection, the "exchange" fibres could be either a second pair of fibres from the break-out tray T (direct exchange fibres), or a pair of output fibres from the mats 6 and 7 (indirect exchange fibres).

These two spare fibres and the associated "exchange" fibres thus constitute a single circuit, this circuit starting at the splice tray 4 which originally stored the two -sl I L I spare fibres, and terminating either at the break-out tray T or at the mats 6 and 7, In this case, the single circuit does not extend to the transmitter/receiver pair of the customer, as the two pairs of the four-fibre unit are housed in the same tube.

The single circuit thus extends between the point in the splice tray 4 where the two pairs of fibres are separated a first fibre separation point), that is to say where a multiple circuit becomes two single circuits and either the break-out tray T or at the mats 6 and 7 (a second fibre separation point). Throughout the length of this single circuit, its fibres are routed separately from the fibres of other circuits, so single circuit management results. Thus, from the point in the first tray 4 where the two pairs of fibres from the customer are separated, the "spare" fibres are routed separately from the two "live" fibres within the tray. On leaving that tray 4, the "spare" fibres pass to another splice tray 4 within an associated bend limiting tube 27b. These fibres are spliced to "exchange" fibres in this second tray 4, and the two "exchange" fibres are passed either to the break-out tray T within a bend limiting tube 27a, or to the mats 6 and 7 within a bend limiting tube 27c. In particular, in the region where re-entry for installation/maintenance purposes to be expected (that is to say in the region of the node single circuit management is ensured, so that optical signals carried by the single circuit are not degraded by installation/maintenance operations carried out on other circuits.

As each of the splice trays 4 is assock!.ed with a respective single circuit, these splice trays are termed single circuit splice trays.

The bend limiting tubes 27a, 27b and 27c of each of the splice trays 4 are provided with respective support manifolds M (see Figures 6 and Each manifold M is a sliding friction fit on a flanged portion (not shown) of the chassis back plate 5c, and is provided with an open aperture 28a for supporting the associated bend limiting tube 27a, and with a pair of closed apertures 28b and 28c for supporting respectively the associated bend limiting tube 27b (if there is one) and the associated bend limiting tube 27c. The AMENDED SHEET I I I I W( (1/071H0 mTAY/44i/i 960 19 manifolds M are made of injection moulded filled nylon.

Figure 10 shows the input mat 6 of the sub-assembly S.

The output mat 7 of this sub-assembly, being of identical construction to the input mat 6, will not be described in detail. The mat 6 includes an input slot 29 for receiving the main input fibre, and an input slot 30 for receiving the standby input fibre. These two slots 29 and 30 lead to an aperture 31 which houses a 2x2 fused coupler (not shown).

The two output fibres from this fused coupler are led via a curved channel 32 around a mandrel 33. The mandrel 33 has a radius of 30mm, and so fulfils the minimum bend requirements for live fibre. Each of the fused coupler output fibres is spliced to an input fibre to a respective 1x4 planar coupler (splitter). The two splices are stored in a recess The two planar couplers (not shown) are housed in an aperture 34 adjacent to the aperture 31. The two fibres pass from the mandrel 33 to their planar couplers via the curved end wall 35a of a recessed portion 35 of the mat 6, and via respective curved slot 36. The eight output fibres of the two planar couplers pass round the mandrel 33 via a slot 37.

These fibres then leave the mat 6 via respective output slots 38 which fan out over the recessed portion 35 and the adjacent raised portion which defines the curved end wall The mat 6 thus forms a 2x8 splitter for the transmit fibres, with one of its inputs being the main transmit input fibre and the other the standby transmit input fibre. As mentioned above, only main fibres are used in normal operation, so the mat 6 acts as a 1x8 splitter. However, should there be problems with the main fibre route, the mat 6 will still act as a 1x8 splitter with the standby receive fibre as its input fibre.

Similarly, the mat 7 constitutes a 2x8 splitter for the receive fibres.

Figure 11 shows the splitter array back cover 8 of the sub-assembly S, in greater detail. The back cover 8 is formed with a pair of longitudinally-extending grooves 8a adjacent i -rrr I -T Usl-~4~ IY~ I~JI*ICIIl~ lsul~~ WO 91/0.106 KI "'179114/11 I to that end remote from the groove 26. These grooves 8a reduce the thickness of the back cover in this end region, and so enhance the flexibility of the back cover, thereby ensuring that, in use, the back cover holds the mats 6 and 7 firmly against the chassis back plate 5c. In this connection, it should be noted that this end region of the back cover 8 is formed with an in-turned L shaped flange 8b which can be snapped over grooves 28d formed in the manifolds M to hold the back cover to the chassis 5 with the mats 6 and 7 firmly sandwiched therebetween.

The outer surface of the back cover 8 is also provided with a plurality of longitudinally-extending ribs 8c, the base of each rib being formed with a plurality of apertures 8d. These apertures 8d extend right through the back cover 8 to its inside surface, and constitute a matrix of tie points for the attachment of cable ties which are used to tie the bend limiting tubes 27a, 27b and 27c to the sub-assembly

S.

Figure 12 shows the break-out tray T in greater detail. As mentioned above, two loops of the cable are stored in the track 3, before the cable exits the break-out tray T via the entry portion 2, and one of the tubes of the cable is cut in the middle of its stored length. One of the cut fibres forms the main fibre for the splitter array subassembly shown in Figures 5 and 6, and another the standby fibre for that sub-assembly. The remaining fibres can be main and standby fibres for other splitter array subassemblies S, and S 3 of the node N, or can be stored around a mandrel 39 at that end of the tray T remote from the cable entry portion 2. The mandrel 39 has a rounded rectangular cro'is-section, and is sized so that fibre coiled therearound does not exceed minimum bend radius requirements.

The break-out region B of the tray T is formed with a plurality of curved upstanding fingers 40, adjacent pairs of which define sixteen fibre feed channels 41. The two fibre end portions that constitute the main fibres associated with the lowest splice tray 4 of the sub-assembly S1, are fed i i 1 I through the first channel 41 (that is to say through the channel nearest the entry portion Similarly, the two fibre end portions that constitute the stanidby fibres associated with the second lowest splice tray 4 are fed through the second channel 41. (As there are sixteen channels 41, the break-out tray T can handle sixteen pairs of fibre end portions, that is to say all the fibre end portions from two cut tubes.) The two fibres then pass into the bend limiting tube 27a associated with the lowest splice tray 4 of the sub-assembly S1. This tube 27a passes through an aperture 42 in a raised portion 43 of the break-out region B (see Figure 13), and is tied in place by ties (not shown) associated with a further aperture 44.

A preferred form of TPON includes a 32-way split, that is to say each fibre from the exchange serves 32 actual customers via one or more splitting (flexibility) points such as the node N described above. As the node N defines an 8-way split, it could be used as a primary splitting point, in which case each of the "customer" fibres leaving the node would lead to a respective secondary splitting point. In this case, each of the single circuits associated with a customer tray 4 would start at the output of the splitter mats 6 and 7 a first fibre separation point), and terminate at the input to the secondary splitting point (a second fibre separation point). Alternatively, if the "customer" fibres leading to the secondary splitting point are fed back into a multi-fibre tube, the single circuit terminates at the entry to this tube (a second fibre separation point). In either case, in the region where re-entry for installation/maintenance purpose is to be expected, single circuit management is ensured.

Each of the secondary splitting points would be similar to the node N, but each incoming fibre would be split four ways rather than eight ways. As the outgoing fibres from a primary node do not go directly to customers, the terms "customer splice trays" and "customer fibres" used above should be taken to mean splice trays and fibres associated with either actual customers or with downstream splitting points. Of course, in the preferred 32-way split AMENDED SHEET ill I I \WO /Wlfl t" II 10/01 (i form of TPON, the nodes N could also be secondary nodes. In this case, there would be four nodes N, each serving eight actual customers, and the four secondary nodes would be served via a 4-way split primary node. Here again, the primary node, would be similar to the node N, but each incoming (exchange) fibre would be split four ways rather than eight ways.

It will be apparent that the type of splitter array sub-assembly described above is extremely flexible in that it can readily be adapted to suit different requirements. In particular, it is adaptable to any splitting ratio by varying the number of splice trays used and the size and form of the splitter array mats 6 and 7. Moreover, by co-locating several splitter array sub-assemblies in a node, splitting from a plurality of exchange fibres can be accomplished at any gi-ren point, using different splitting ratios in each sub-assembly if required.

It will be apparent that the arrangement described above ensures single circuit management. In particular, in the region of a node N (that is to say where re-entry for Installation/maintenance purposes is to be expected) single circuit management is ensured, so that optical signals carried by any single circuit are not degraded by installation/maintenance operations carried out on any other circuit.

An important advantage of the sub-assemblies described above, is that the splitters and the associated fibres can all be factory fitted. Thus, the fused and planar couplers and their associated fibres can be made and positioned in the mats 6 and 7, and the associated fibres can be led to their splice trays 4 within bend limiting tubes all at the factory. When the sub-assembly is to commissioned, the fitter needs only to cut one or more tubes of the cable C, feed main and standby fibre end portions to the lowest two splice trays 4 of the sub-assembly, store spare cut fibre end portions in the break-out tray T, splice the main and standby fibre end portions to the main and standby input fibres I ul JI I I- 11 I .i roady prooont in tho two opli±o trays, and to opliico "customer" fibres to the fibres already present in the other splice trays 4, In this way, the amount of skilled work which has to be carried out on s e is reduced to a minimum.

n particular, the fitter does not need to carry out any Intricate splicing for splitting purposes. Moreover, the bend limiting tubes ensure that the fibres guided therein are never bent beyond the minimum bend requirements for live fibre even when the splice trays are pivoted out of the stack to either of their operating positions. This guarantees the transmission performance of live fibres carried by the bend limiting tubes.

The sub-assembly described above could also be adapted for use in a spur joint. In such a case, no splitting is required, so the sub-assembly would not include the mats 6 and 7. In a first type of spur joint, all twelve tubes of the fibre cable C would be cut, thereby forming twelve main fibre tube ends and twelve standby fibre tube ends. The fibres of six of the main fibre tube ends would then be spliced to the fibres of six of the standby fibre tube ends in single element splice trays 45 (as is described below with reference to Figure 14). The fibres of remaining six main fibre tube ends are then spliced to "customer" fibres in 24 single circuit splice trays 4. Similarly, the fibres of the remaining six standby fibre tube ends are spliced to 48 "customer" fibres in 24 single circuit splice trays 4. Thus, two fibres are fed, in bend limiting tubes, from a breakout tray (not shown) to each of the 48 single circuit splice trays 4, where they are spliced to "customer" fibres in a similar manner to that described above with reference to Figures 5 and 6. As with the customer splice trays 4 discussed above, single circuit management results for each of these 48 splice trays.

A respective main fibre tube nd and a respective standby fibre tube end are fed from the break-out tray to each of the single element splice trays 45 (see Figure 14), each tube end being in a respective bend limiting tube (not

I

~gaLIR~gMsn~a~~m WOIOW b/liP 14 "IM, 119411119611 :!hown, but 9imilar to tho bend limiting tubes 27a, 27b and 27c). Each tray 45 has a main body portion 46 and a tube entry portion 47. The main body portion 46 is of oval configuration, having an oval base 46a and an upstanding peripheral wall 46b. Fibre access to the main body portion 46 from the tube entry portion 47 is via a channel 48.

Channels 49, 50, 51 and 52 are provided in the main body prortion 46 to guide both main and standby fibres to a pair of splice holder regions 53. The channel 51 is a direction reversing channel, and permits main and standby fibres to approach each of the splice holder regions 53 from opposite directions.

Each single element splice tray 45 is also provided with a number of fibre retention tabs 54 for holding fibre in the various channels 49 to 52.

The tube entry portion 47 of each single element splice tray 45 includes two tube entry/exit ports 55a and Channels 56a and 56b are provided to lead fibre between the ports 55a and 55b and the channel 48.

The single element splice tray 45 is provided with a leash 57 and a retaining ring 58 (similar to the leash 24 and the retaining ring 25 of the tray The leash 57 permits the tray 45 to be pivoted out of a stack of trays to enable ac '3s to the tube entry portion 47.

In use, a main fibre tube end is led to the port of each of the splice trays 45, and a standby fibre tube end is led to the port 55b of each of the splice trays Inside each tube entry portion 47, the tubes are cut away to expose the fibres. The fibres are then fed into the main body portions 46 of the trays, where they are spliced. The eight splices in each tray 45 are then positioned, four in each of a pair of splice holders, and the splice holders are then mounted in the regions 53. The fibres leading to the splices are then stored in the main body portions 46 of the trays 45. A length of each of the fibres is stored in the main body portion 46 of the associated tray 45 by passing these fibres one or more times round an upstanding mandrel 59 r and under the tabs 54, The fibres' natural resilience will ensure that the loops of fibre expand outwardly to a configuration of varying diameter turns, The provision of stored fibre permits re-splicing to be carried out during the lifetime of the assembly.

In a modified version of the spur joint described above, only six of the tubes are cut, the fibres in these tubes being spliced to "customer" fibres in 48 single circuit splice trays 4 as described above. The remaining six uncut tubes are looped around a break-out tray. Alternatively, instead of using 48 single circuit splice trays 4, six single element splice trays 45 could be used. This alternative would, however, only be used in cases where there is no need to access spur joints for future use.

Obviously, for either type of spur joint, the number of fibres forming the spur can be varied. For example, the spur could be formed from the fibres of one cut tube. In this case the spur would contain 16 fibres (eight main fibres and eight standby fibres from a single cut tube) and 88 fibres (from the remaining eleven tubes either cut and spliced or uncut and coiled) would continue through on the ring. In each case, however, single circuit management is ensured for each of the single element splice trays 45. Thus, the single circuit associated with a given single element splice tray 45 starts and terminates at the break-out tray T (first and second fibre separation points), a respective main fibre tube end being fed to the splice tray 45 in a bend limiting tube, where its fibres are spliced to the fibres of the standby fibre tube and which, in turn, is fed back to the break-out tray within a bend limiting tube. The single circuit here is, therefore, constituted by a plurality of fibres (eight in the example above).

AMENDED SHEET

Claims (2)

1. 26- THE CLAIMS DEFINING THE INVENTION ARE AS FOLLOWS:- 1. A method of providing single circuit management within a network node of a passive optical network, the method including the steps of: routing input optical fibre(s) from a cable element to an input splice tray; (ii) splicing said input optical fibre(s) to optical splitter input fibre(s) in said input splice tray, said optical splitter input fibre(s) being optically coupled to an optical splitter, said optical splitter being optically coupled to a plurality of optical splitter output fibres and each of the optical splitter output fibres being spliced to a respective output optical fibre within a respective output splice tray; and (iii) routing each of the output optical fibres from their respective output splice tray to a respective output optical fibre cable or a respective output optical fibre cable element such that the input optical fibre(s), the splitter input fibre(s), the splitter output fibres and the output optical fibres are not bent to a radius less than the minimum bend radius of live optical fibre, wherein the optical fibre(s) forming each single circuit is separately routed through a dedicated input splice tray and a dedicated output splice tray such that the optical fibre(s) forming each single circuit can be accessed within said dedicated input splice tray or said dedicated output splice tray without causing substantial disturbance to the optical fibre(s) forming another single circuit. 2. A method according to claim 1, wherein the input optical fibre is routed from the cable element to the input splice tray via a bend limiting tube. 3. A method according to claim 1 or claim 2, wherein each output optical fibre is routed from the output splice tray to the respective output optical fibre cable or the respective output optical fibre cable element via a respective bend limiting tube. 4. A method of providing single circuit management within a network node of an optical network, the method including the steps of: routing an input optical fibre from a cable element to a splice tray; (ii) splicing said input optical fibre to an output optical fibre within said splice tray; and (iii) routing said output optical fibre from said splice tray to an output optical fibre cable or an output optical fibre cable element such that the input
2. 27- optical fibre and the output optical fibre are not bent to a radius less than the minimum bend radius of live optical fibre, wherein the optical fibre(s) forming each single circuit is separately routed through the splice tray such that the optical fibre(s) forming each single circuit can be accessed within said splice tray without causing substantial disturbance to the optical fibre(s) forming another single circuit. A method according to claim 4, wherein the input optical fibre is routed from the cable element to the splice tray via a bend limiting tube. 6. A method according to claim 4 or claim 5, wherein the output optical fibre is routed from the splice tray to output optical fibre cable or the output optical fibre cable element via a bend limiting tube. 7. An optical fibre management apparatus for use in a node of an optical fibre network comprising a cable element storage area, one or more input splice trays, one or more optical splitters and one or more output splice trays, the optical fibre management apparatus being configured to: guide an input optical fibre from a first cable element stored in the cable element storage area to one of the input splice trays; (ii) guide an optical splitter input fibre from said input splice tray to one of the optical splitters; 2 (iii) guide an optical splitter output fibre from said optical splitter to one of the output splice trays; and (iv) guide an output optical fibre from said output splice tray to a second cable element stored in the cable element storage area such that the input optical fibre, the optical splitter input fibre, the optical splitter output fibre, the output optical fibre, the first cable element and the second cable element are managed such that they are not bent to a radius less than the minimum bend radius of live optical fibre. 8. Optical fibre management apparatus according to claim 7 additionally comprising at least one length of bend-limiting tubing, the or each length of bend- limiting tubing guiding at least one of the input optical fibre, the optical splitter input fibre, the optical splitter output fibre and the output optical fibre such that they are not bent to a radius less than the minimum bend radius of live optical fibre. -28- 9. Optical fibre management apparatus according to claim 1 or claim 8, wherein the one or more input splice trays and the one or more output splice trays are arranged in a vertical stack. An optical fibre management apparatus substantially as hereinbefore described with reference to any one of the disclosed embodiments and its associated drawings. 11. A method of providing single circuit management within a network node of an optical network substantially as hereinbefore described with reference to any one of the disclosed embodiments and its associated drawings. DATED this 17th day of July, 1998 BRITISH TELECOMMUNICATIONS public limited company Attorney: PETER R. HEATHCOTE Fellow Institute of Patent Attorneys of Australia of BALDWIN SHELSTON WATERS