GB2315626A - Amplified optical network - Google Patents

Abstract

An optical network in which there is upstream communication of traffic from a set of outstations (11) to a basestation (10) via a net of transmission pathways that includes an active optical combiner (26). This combiner has a set of n input ports (30) coupled to a single output port (34) via the series combination of a single opto-elecronic transducer (31), a electrical amplifier (32), and an electro-optic transducer (33). The coupling from the input ports to the opto-electronic transducer is multimode in order to avoid the combining loss of a single mode combiner. If short enough, the individual transmission paths from the outstations to the combiner may also be multimode, with a consequent potential for saving in outstation transmitter cost. The input fibres may be tapered and arranged in a hexagonal matrix. A similar, parallel, downstream communications arrangement is also provided.

Description

OPTICAL NETWORKS This invention relates to optical networks in which there is optical communication to a basestation from a plurality of outstations. Typically those networks providing two-way communication between the basestation and the outstations. In a basic form of such a network there is no intermediate amplification of signals travelling in either direction between the basestation and any of the outstations, and such networks are commonly referred to as Passive Optical Networks or PONS. Figure 1 depicts the basic structure of a PON having a single basestation 10 and a set of n outstations 11 (only three illustrated). Downstream traffic from the basestation to the outstations is transmitted down a single physical transmission pathway 12 to an n-way splitter 13 which divides the downstream signal substantially equally between n branches 14. A different one of each of the n branches 14 is connected to a different one of each of the n outstations. Upstream traffic from the n outstations is fed through individual branches 15 to an n-way combiner 16 for onward transmission to the basestation 10 via a single physical transmission pathway 17. The n-way splitting does not have to take place at a single location, but may be organised in the form of a tree Inot shown) of sub-unit splitters at locations spaced from one another. A similar situation exists also in respect of the n-way combining. The data transmission is typically based on the Time Division Multiplex/Time Division Multiple Access (TDM/TDMA) protocol. In the downstream direction, information is transmitted in a TDM frame with data for a particular outstation being recognised and extracted on the basis of its position in the frame. In the upstream direction outstations at different propagation distances from the basestation will have different transmission delays. The TDMA protocol prevents collision of information from different outstations by arranging for bursts of data to be transmitted from each outstation at a timing adjusted so that they will arrive at the basestation in the correct timing relationship with respect to bursts from other outstations, thereby effectively forming a TDM frame at the basestation.

In the downstream direction it is quite clear that an individual outstation can receive only a fraction of the power transmitted by the basestation because the basestation power is shared between all the outstations.

This means that the basestation transmitter must be more powerful than would be the case for an equivalent single point to single point transmission. A similar situation pertains in respect of the upstream direction, not because the transmitted power of an outstation transmitter has to be shared, but because the distance between the basestation and the outstations is usually so great that it is necessary to use single mode transmission, and it is not possible to combine two single mode signals of the same frequency and polarisation state into a single mode signal without losing at least 3dB of power.

It is known that the range of a PON can be extended by incorporating intermediate amplification into the transmission paths between basestation and outstations, as for instance described in a paper by J A Quayle entitled 'Ranging on Advanced PONs' presented at the Thirteenth Annual Conference on European Fibre Optic Communications and Networks, Brighton, England 1995. This paper is particularly concerned with the implications of using intermediate amplification so far as the provision of ranging information for separating the times of arrival at the basestation of data received from the different outstations.

The present invention is similarly concerned with an optical network in which single basestations are optically coupled with pluralities of outstations via optical transmission paths that include intermediate amplification, and is particularly concerned with reducing the power requirements of the outstation transmitter of such systems.

According to the present invention there is provided an optical network in which there is upstream communication of traffic from a plurality of outstations to a basestation via a net of optical transmission pathways that includes an n-way optical combiner having n input ports coupled to an output port via the series combination of an opto-electronic transducer an electrical amplifier and an electro-optic transducer, wherein the coupling between the input ports and the opto-electronic transducer is multimode.

The invention makes use of the fact that an opto-electronic transducer such as a photodiode is not intrinsically a single mode device, and that therefore the losses associated with single mode combining of optical power at an n-way optical combiner can be avoided, provided that the combiner includes opto-electro-optical conversion and amplification, by combining the transmissions from a set of outstations in a multimode fashion upon the photosensitive surface of an opto-electrical transducer such as a photodetector. The avoidance of this source of power loss may be used to effect savings elsewhere in the network. For instance, for a given length of pathway between outstation and combiner, the minimum power output requirement set for the outstation transmitter can be reduced, typically involving a cost-saving. Alternatively, provided that it is not dispersion limited, the length of the pathway between outstation and combiner can be extended, for a given minimum power output level of the basestation transmitter.

Resulting from the fact that all of the light received from a basestation is incident upon the photodetector, there is no need to prevent light from more than one basestation being simultaneously incident upon the photodetector because, even if this should give rise to coherent interference effects, this will not give rise to modal noise in the electrical output of the photodetector. A change in relative phase between two or more signals close enough in frequency to be coherently interfering will cause spatial modulation of the interference pattern formed on the photosensitive surface of the photodetector but, since the photodetector collects all the light, no mode selective loss mechanism is present and hence the linearity of response of the photodetector ensures that the spatial modulation is not converted to electrical modulation of the output of the photodetector. This means that the base stations can be safely operated, if desired, with injection lasers energised on a cw basis in order to avoid the bandwidth limitations that are otherwise imposed by the delay associated with having such a laser on from beneath lasing threshold.

There follows a description of optical networks embodying the present invention in preferred forms. The description refers to the accompanying drawings in which: Figure 1(to which previous reference has been made) is a schematic diagram of a PON, Figure 2 is a schematic diagram of an optical network differing from that of the PON of Figure 1 by the use of amplifiers in the optical combiners and splitters of the upstream and downstream parts of the transmission paths between basestation and outstations, Figure 3 is a schematic block diagram of the optical combiner of the network of Figure 2, and Figures 4, 5 and 6 are schematic diagrams of these different ways by which light from outstations of the network of Figure 2 may be directed on to the photodetector of the combiner unit of Figure 3.

The optical network of Figure 2 is distinguished from the PON of Figure 1 in that its optical splitter 25 and optical combiner 26 are active devices that incorporate amplifiers that permit the network a greater range than that of the PON of Figure 1. The basic components of the optical combiner 26 are depicted in Figure 3 and comprise a set of n optical input ports 30 optically coupled with an opto-electronic transducer 31, typically a photodiode, an electrical amplifier 32, an electro-optic transducer 33, typically a semiconductor laser, and a single output port 34 optically coupled with the output of the laser 33. The transmission paths 15 from the outstations 11 to the combiner 26 are constituted by optical fibres, and so at the combiner these fibres can be brought together and arranged to terminate in a close-packed array to constitute the input ports 30 of the combiner. If the photosensitive area of the opto-electronic transducer 341 is large enough, the end of the closepacked array of fibres could be directly imaged upon the photosensitive area or directly butted up against it. However, if the fibres are single mode fibres, this approach does not make efficient use of the photosensitive area because the spot size of light guided by such a single mode fibre is so much smaller than the cross-sectioned area of the fibre that the effective packing fraction of the fibre bundle, in illumination terms, is small.

One way of improving upon this situation is illustrated in Figure 4. In this arrangement light is guided by optical fibres 41 from the ports 30 (not shown in Figure 4) most of the way to the photosensitive area 40 of the opto-electronic transducer 31.

From the input port ends of these fibres, the fibres are gathered together into a close-packed bundle, and then, in the close-packed bundle, the final part of each fibre is uniformly and adiabatically tapered, thereby forming the tapered bundle depicted at 42 in Figure 4. To make such a taper, a bundle of fibres created by forming a hexagonally close-packed assembly of n fibres that are fixed together over an intermediate portion of their length where they are in the close-packed hexagonal array. The assembly is then drawn down in a controlled manner to form two tapers (not shown) joined by their smaller ends. Preferably these tapers are made by the progressive stretching technique described in GB 2 150 703 in which the bundle is longitudinally traversed several times through a localised hot zone using two translation stages, the leading one of which is moved at a controlled rate faster than the trailing one so as to produce plastic flow strain in the bundle where it is locally softened by the heat of the localised hot zone. By this means two adiabatic tapers are formed whose smaller ends are linked by a parallel-sided region of the bundle of reduced cross-section. The schedule of progressive stretching is arranged to provide these two tapers with a chosen taper half-angle, typically a half-angle lying between 10 and 2", so that, when the reduced cross-sectional area parallel-sided region is removed, all the fibres of a bundle taper 42 (Figure 4) point to a common point a short distance in front of the small end of the taper. The taper 42 is positioned in relation to the opto-electronic transducer 31 so that the common point lies at the centre of the photosensitive area 40 of that transducer. The spacing between the small end of the taper 42 and the photosensitive area of the transducer 31 means that, if a hermetic packaged environment is required for the transducer, the taper 42 does not have to penetrate the hermetic package, but instead, light issuing from the small end of the taper may be coupled into the package through a window 43 formed in a well (not shown) of the package (not shown).

In the introduction to the description relating specifically to Figure 4 it was stated that the problem addressed by the particular configuration of coupling between input ports and opto-electronic transducer of the optical combiner was one arising from the small packing fraction represented by the optical cores of a bundle of close-packed single mode fibres. It was also explained that single mode fibres are typically employed throughout a PON because the transmission distances are too great to permit satisfactory use of multimode fibre. On the other hand, in the case of optical networks according to the present invention that incorporate optical amplification in their optical combiners, if the transmission paths 15 between basestations and the optical combiner are short enough, for instance because the basestations are all in a common building, multimode fibre may be used for those transmission paths even though the transmission path 17 from combiner to basestation may be significantly longer, and so requires the use of single mode fibre. Under these circumstances there can be a useful cost saving in the optical sources employed in the outstations, for instance by the use of non-coherent semiconductor diodes instead of diode lasers, or reducing laser packaging costs through being able to accept less stringent alignment tolerances through the use of the better light collection efficiency of multimode fibres. If the transmission path fibres 15 are multimode, then the fibres 41 may also be multimode. The larger core packing fraction of the fibre bundle at the combiner 26 then simplifies the coupling to its opto-electronic transducer 31. The electrical output of this transducer, once amplified, is suitable for drawing a single mode output electro-optic transducer 33 for the onward transmission by single mode fibre 17 to the basestation 10, this notwithstanding that the optical input to the opto-electronic transducer 31 is a multimode input Reverting attention to systems employing single mode fibre transmission pathways 15 from the outstations 11 of Figure 2 to the combiner 26, Figures 5 and 6 depict further alternative forms of coupling between the input ports 30 of the combiner and its opto-electronic transducer 31.

Like the form of coupling of Figure 4, the couplings of Figures 5 and 6, both employ tapered close-packed fibre bundles, but the tapered bundles 52 of Figures 5 and 6 are different from the tapered bundle 42 of Figure 4. In the case of the tapered bundle 42, the tapered angle is chosen so that all the fibres shall point to a common point at the centre of the photosensitive area 40 of the opto-electronic transducer 31. In the case of the tapered bundles 52, the taper angle is not critical provided that it is kept within adiabatic limits, and instead the taper is employed to reduce the cross-sectioned area of its constituent fibres so as to produce an increase in modal spot size and an attendant reduction in beam divergence of light issuing from the small end of the fibre. In the case of conventional 125pom diameter single mode transmission fibre, an adiabatic taper that reduces the diameter down to about 90pom increased the modal spot size from about 10pom to 15pom.

If the photosensitive area 40 is large enough, the small end of the taper 52 may be brought into abutting relationship, as depicted in Figure 5. If, however, the opto-electronic transducer needs to be hermetically packaged, the difficulty of providing a hermetic feed-through of a bundle of fibres through the package wall can be avoided, as depicted in Figure 6, by coupling the small end of the taper 52 to the photosensitive area 40 via a single short length 60 of multimode fibre hermetically sealed through the package wall (not shown). The inboard end of the multimode fibre 60 is shown as butted against the photosensitive area, but it could alternatively be optically coupled by means of a lens (not shown). Such a lens could also be used in place of the multimode fibre 60, and in the case of a hermetically packaged opto-electronic transducer 31, such a lens can be accommodated externally or internally of a window in the package wall, or can itself form a part of the package wall.

Claims (5)

CLAIMS:

1. An optical network in which there is upstream communication of traffic from a plurality of outstations to a basestation via a net of optical transmission pathways that includes an n-way optical combiner having n input ports coupled to an output port via the series combination of an opto-electronic transducer an electrical amplifier and an electro-optic transducer, wherein the coupling between the input ports and the opto-electronic transducer is multimode.

2. An optical network as claimed in claim 1, wherein there is additionally provision for downstream communication of traffic from the basestation to the plurality of outstations.

3. An optical network as claimed in claim 1 or 2, wherein a tapered close packed optical fibre bundle is included in the coupling between the n input ports and single opto-electronic transducer of the combiner.

4. An optical network as claimed in any preceding claim, wherein the optical transmission pathway between the outstations and the n-way optical combiner is a multimode pathway.

5. An optical network substantially as hereinbefore described with reference to Figures 2 to 6 of the accompanying drawings.