WO2005106550A1

WO2005106550A1 - Dual band optical mux/demux and apparatus comprising the same

Info

Publication number: WO2005106550A1
Application number: PCT/EP2004/004583
Authority: WO
Inventors: Matteo Cherchi
Original assignee: Pirelli & C. S.P.A.
Priority date: 2004-04-30
Filing date: 2004-04-30
Publication date: 2005-11-10

Abstract

An optical device for splitting/combining a first and a second optical wavelength band has a first, a second and a third optical splitting device optically coupled in cascade and a first and a second differential optical delay device optically coupled to, and interleaved between, the optical splitting devices. This structure together with a suitable choice of the coupling coefficients of the splitting devices and of the differential delays of the differential optical delay devices achieves flattened pass bands and stop-bands and makes the optical device tolerant fluctuations of the structural parameters. An apparatus comprising the optical device is also disclosed for use in fiber-to-the-premises networks.

Description

DUAL BAND OPTICAL MUX/DEMUX AND APPARATUS COMPRISING THE SAME

* * * * * DESCRIPTION The present invention generally relates to the field of optical wavelength splitters/combiners, more particularly it relates to fiber-to-the-premises network equipments comprising optical wavelength band splitter/combiner. Currently deployed fiber-to-the-premises (FTTP) networks incorporate gigabit passive optical network (GPON) for broadband delivery of voice, video and high- speed data directly to the home or broader community through optical fiber. Converged voice, video and data services networks are also known as "triple play networks". These networks support two signals in downstream direction (from central station to the user) and one signal in upstream direction. A first downstream signal delivers analog television and a second downstream signal delivers digital voice and data services, such as for example telephone and/or Internet. The upstream signal is typically a digital signal delivering voice and data from the user to the central station or the service provider. Typically, FTTP system delivers voice, video and data over a PON using the ITU-T GPON standard. The system supports radio frequency (RF) analog video delivery using a 1550 nm wavelength overlay. High-quality video on a PON is achieved with a high power 1550 nm signal and power requirement at 1550 nm is greatly demanding. The second downstream signal uses a 1490 nm wavelength and the upstream digital signal is typically allocated at 1310 nm wavelength. In FTTP networks, as well in many other applications, a key technology is signal splitting and/or combining. Signal multi/demultiplexing must fulfill very demanding requirements such as, among the other, wide bandwidths and small crosstalk over a wide temperature range (typically from -40°C to +85°C). In fact, low cost components, such as light emitter, are used over a wide temperature range, possibly generating a large wavelength drift. Many applications require a rectangular wavelength response in order to maintain a low-loss and wavelength- independent transmission in a passband and a high-level rejection to all wavelengths in a stopband. For example, anticipated telecommunications applications seek a 1.3/1.55μm WDM filter having a flat and low-loss passband at 1.280-1.335 μm and a -50 dB stopband at 1.525-1.575 μm. Various devices have been proposed to fill these new, demanding requirements but none are fully satisfactory.

When the separation of two wide bands widely spaced (centered, for example, around λ^lSOOnm and λ₂=1500nm) is needed, micro-optic products based on thin-film technology are currently used, such as for example multilayer thin-film filters in free space optics. Nonetheless, they are undesirable because they cannot be readily integrated and because their fabrication requires high labor cost, mainly due to the operations of alignment of components during assembly, and because of difficulties in coupling light to and from fibers. Optical splitting components may be classified as the following three types: (1) bulk-type optical splitters; (2) fiber-type optical splitters; and (3) guided-wave type optical splitters.

The bulk-type optical splitters are constructed by arranging microlenses, prisms, interference-film filters, etc., and have little wavelength dependence. Although the bulk-type optical splitters can be put into practical use to some extent, they require a long time and skill for assembly and adjustment, and present some problems with regard to long-term reliability, cost and size.

The fiber-type optical splitters are fabricated using optical fibers as constituent material. Although they exhibit reduced wavelength dependence, the fabrication process requires skill, and is not suitable for mass production because of lack of reproducibility.

In contrast, guided-wave type optical splitters have the advantage that they can be constructed on flat substrates in large quantities through processes such as the photolithography process. Hence, they attract attention as a promising type of splitting component which can be reproduced and integrated as compact parts. Optical devices based on this technology are also referred to as planar lightwave circuits (PLC) devices or integrated optical circuit (IOC) devices. FIG. 1 is a planar view exemplifying a configuration of a conventional (2x2) guided-wave type optical coupler. In FIG. 1 , two optical waveguides 2 and 3 are formed on a substrate 1. A part of the optical waveguide 2 and a part of the optical waveguide 3 are brought into close proximity with each other over a length L_c to form a directional coupler 4. One end of the optical waveguide 2 is a first input port 7 into which an optical signal P_in is launched, and the other end of the optical waveguide 2 is a bar output port 8 from which a bar optical signal Pba_r is emitted. Similarly, one end of the optical waveguide 3 is a second input port 5, and the other end of the optical waveguide 3 is a cross output port 6 from which a cross signal P_croSs is emitted. The directional coupler 4 may be designed in such a way that an optical signal P_in launched into the first input port 7 is branched into two optical signals P _ar and P_cr0Ss to be outputted from ports 8 and 6, respectively. At a given wavelength, the power coupling ratio C is defined as:

when the input optical radiation is launched only in the first input port 7. Generally, the power coupling ratio C of an optical coupler of the type of figure 1 may be expressed by the following equation: C=sin² θ(λ) (2), wherein the angular coupling coefficient θ(λ) of the power coupling ratio typically depends on wavelength. In case of a directional coupler 4 of the type of figure 1 , θ(λ) also depends on the length L_c of the straight coupling region, on the distance between the waveguides at the coupling region over the length L₀, on the shape, width and depth of the waveguides, on the refractive index difference between waveguide core and cladding, on the geometry of the input and output curves, etc. The optical coupler 4 can act as an optical wavelength splitter, by properly exploiting the wavelength dependence of the power coupling ratio, but its sine-like response doesn't make it suitable for telecommunication purposes. On the other hand, although the power coupling ratio of the directional coupler 4 can be specified to a desired value at a particular desired wavelength, the wavelength dependence of the coupling ratio presents a problem when the optical coupler is used in a wide wavelength region. Mach-Zehnder (MZ) interferometers have been widely employed as optical band splitter/combiner, but they have a sinusoidal response, giving rise to strongly wavelength-dependent transmission and a narrow rejection band. Other designs have encountered a variety of practical problems. The article of M. Cherchi "Wavelength-flattened directional couplers: a geometrical approach" Applied Optics, Vol. 42, N° 36, pp.7141-7148, Dec. 2003 shows that the action of a generic 2-branch optical device may be decomposed in the product of a differential delay φ between the two branches and a coupling θ between the two branches. Figure 2 shows a geometrical representation of the action of a 2-branch optical device at a given wavelength in the particular case of a pure differential delay. In Figure 2, a generic point P on a spherical surface S represents a generic normalized mode of a 2-branch optical device. Points E₁ and E₂ along axis S₁ represents the modes wherein the optical power is, respectively, totally in the first and in the second optical branch of the 2-branch optical device. Points E_A and E_s along axis S₂ represents respectively the antisymmetric and the symmetric modes of the two optical-branch system. In case of a pure differential delay φ (i.e. without coupling), the point P representing a generic input state will be rotated by an angle φ on the circle of revolution on S about axis Si and passing through P, represented by curve 10. Conventionally, a counter-clockwise revolution, when the observer is along axis Si, will be associated to a positive differential delay φ, as shown in figure 2.

Similarly, figure 3 shows a geometrical representation of the action of a pure (i.e. without optical delay) 2-branch optical coupler having coupling coefficient θ. The point P representing the generic input state will be rotated by an angle 2Θ on the circle of revolution on S about axis S₂ and passing through P, represented by curve 20. Conventionally, a clockwise revolution, when the observer is along axis S₂, will be associated to a (positive) coupling action, as shown in figure 3. Article K. Jinguji, N. Takato, Y. Hida, T. Kitoh, and M. Kawachi, "Two-port optical wavelength circuits composed of cascaded Mach-Zehnder interferometers with point-symmetrical configurations," J. Lightwave Technol., vol. 14, pp. 2301-2310, Oct. 1996, discloses a wavelength multi/demultiplexer with a flattened passband and stopband. It has a three-stage MZI configuration with a point-symmetry arrangement. The four directional couplers comprised in the multi/demultiplexer are the same. Applicant has determined that the structure disclosed in Jinguji et al. is based on a flattening principle which may result in a limited tolerance to fabrication errors, especially in view of a dense optical integration, as will be described in greater details below. In addition, Applicant has determined that the intrinsic complexity of this structure, due to the presence of four couplers and the corresponding eight S- bends, may give rise to fabrication difficulties, additional loss, especially in the context of dense optical integration, and to additional length for the device, which is of paramount importance in high density IOC.

Monolithic optical waveguide devices are particularly promising because they can perform complex circuit functionalities and because they can be made by mass production integrated circuit techniques. The integration of all the components needed for the full functionality in a single optical integrated circuit may reduce the alignment problem. Moreover, a single integrated chip may allow to a larger extent the automation during the module assembly.

In this context, it is highly desirable to achieve a high-density of optical chips on the substrate wafer, in order to reach an economic cost-effectiveness. In fact, the higher is the density of PLC devices on the substrate, the higher is the saving in the labour-and-material cost and the shorter is the overall processing time. In order to achieve an high-density, it is convenient to choose a high waveguide-to- cladding index contrast, wherein high index contrast means an index contrast equal to or greater than 1%. A problem arises with high index contrast, in that the fabrication tolerances and chip-to-fiber alignment tolerances worsen with the increasing of index contrast. Nonetheless, cost-effectiveness of the PLC device demands that there is little need of active characterization and/or control of the device, both during the fabrication process and during operation. The active characterization and/or control may be avoided if the device is strongly tolerant to fabrication errors, in order to guarantee high fabrication yield, and strongly tolerant to operation variations, such has wavelength drifting and/or temperature variations. Accordingly, the demand for high density and the demand for high tolerances are in conflict. The Applicant has tackled the problem of providing an optical band splitter/combiner device, particularly suitable for FTTP networks, having wide flattened bands and strong tolerance to fabrication errors. The Applicant has found that the above problem is particularly severe in the context of PLC devices, and more particularly in the context of high-density PLC devices. The Applicant has designed a new splitter/combiner of a first and a second optical band, comprising three cascaded splitting devices and two differential phase delay devices alternated with the splitting devices, wherein the three splitting devices are designed so as to send the power input at an ideal operating wavelength λ_1op within the first optical band completely in the respective cross output port and at an ideal operating wavelength λ_2op within the second optical band completely in the respective bar output port, and wherein the phase difference provided by the first differential phase delay device has a value of approximately 120° both within the first optical band and the second optical band and the phase difference provided by the second differential phase delay device has a value of approximately 240° within the first optical band and approximately 120° within the second optical band. This splitter/combiner has particularly wide flattened bands and very strong tolerance to fabrication errors.

Using a geometrical approach, Applicant has found a design scheme which is not point-symmetric and achieve filters with low loss and good extinction over a wide band. Cascaded configurations of the inventive structure can be designed in order to get even better performances.

The present invention finds particularly advantageous application in the context of integrated optics circuit or PLC devices, particularly silica based PLC, wherein the device may comprise two optical waveguides apt to form the couplers and the differential delays. Moreover, Applicant has found that the present invention is particularly needed in the context of high-density circuits, wherein the refractive index contrast is greater or equal to 1%, preferably above 2%. Applicant has found that an optimal choice of the refractive index contrast suitable to meet the above requirements is below 4.5%, preferably below 3%. The present solution shows a good trade-off between high manufacturing automation, high density, high yield and high tolerance.

In a first aspect, the present invention relates to an optical device for splitting/combining a first and a second optical wavelength band, the device comprising a first, a second and a third optical splitting device optically coupled in cascade; a first differential optical delay device optically interposed between the first and second optical splitting device; and a second differential optical delay device optically interposed between the second and third optical splitting device; wherein: each of the first, second and third optical splitting device is substantially a full splitter at least at a respective first wavelength within the first optical band and a null splitter at least at a respective second wavelength within the second optical band; the first differential optical delay device has an associated first net phase difference substantially between about 100° and 140° at least at a third wavelength within the first optical band and substantially between about 100° and 140° at least at a fourth wavelength within the second optical band; and the second differential optical delay device has an associated second net phase difference of between about 220° and 260° at least at a fifth wavelength within the first optical band and between about 100° and 140° at least at a sixth wavelength within the second optical band. Preferably, the first differential optical delay device is a zero order differential optical delay device in the first and second optical band.

Preferably, the second differential optical delay device is a higher order differential optical delay device in the first and second optical band. More preferably, the second differential optical delay device has the same order in the first and in the second optical band, exemplarily it is a first order differential optical delay device.

Advantageously, the optical device is a PLC optical device.

In particular, the optical device may comprise a pair of optical waveguides forming the first, second and third optical splitting device and the first and second differential optical delay device.

The optical waveguides have an index contrast preferably higher than about 1%; moreover, the optical waveguides have an index contrast preferably lower than about 4.5%. In a second aspect, the present invention relates to an optical device for splitting/combining a first and a second optical wavelength band, comprising at least two optical devices as described above, optically connected in cascade.

In a third aspect, the present invention relates to an optical network unit comprising - an optical device for splitting/combining a first and a second optical wavelength band as described above;

- an optical receiver optically connected to the optical device and apt to receive a first signal within one of the first and second optical band; and - an optical transmitter optically connected to the optical device and apt to transmit a second signal within the other of said first and second optical band.

Preferably, the optical device for splitting/combining a first and a second optical wavelength band comprises an input port apt to be optically connected to an optical transmission line apt to propagate said first and second signal.

In a further aspect, the present invention relates to an optical network suitable to operate at least at a first and at a second optical transmission wavelength respectively within a first and a second optical wavelength band, the network comprising at least an optical device as described above for splitting/combining said first and second optical wavelength band.

The features and advantages of the present invention will be made apparent by the following detailed description of some exemplary embodiments thereof, provided merely by way of non-limitative examples, description that will be conducted by making reference to the attached drawings, wherein:

- Figure 1 shows a symbolic diagram of a prior art optical coupler;

- Figure 2 shows a geometrical representation of the effect of an optical differential delay device on the optical operative states;

- Figure 3 shows a geometrical representation of the effect of an optical coupler on the optical operative states;

- Figure 4 shows a schematic diagram of an optical device in accordance with the present invention; - Figure 5 shows a geometrical representation of the effect of an exemplary optical device in accordance with the present invention at a first wavelength;

- Figure 6 shows a geometrical representation of the effect of an exemplary optical device in accordance with the present invention at a second wavelength; - Figure 7 shows a schematic diagram of a preferred embodiment of the present invention;

- Figure 8 shows a cross-sectional view of the device of figure 7 taken along line A-A;

- Figure 9 shows a schematic diagram of an exemplary design of an optical splitter in accordance with the present invention;

- Figure 10 shows a schematic diagram of an exemplary design of a first optical differential delay in accordance with the present invention;

- Figure 11 shows a schematic diagram of an exemplary design of a second optical differential delay in accordance with the present invention; - Figure 12 shows a schematic diagram of an optical device comprising a combination of optical devices as in figure 4 or 7;

- Figure 13 shows simulation results of the spectral response of a particular embodiment of the present invention;

- Figure 14 shows simulation results of the spectral responses of a statistical set of particular embodiments of the present invention; and

- Figure 15 shows a schematic diagram of an optical device for FTTP networks comprising the present invention optical device.

Fig. 4 shows a symbolic diagram of an exemplary optical device 100 according to the present invention. The device 100 comprises a first, a second and a third optical splitting device, represented respectively by blocks 106, 107 and 108, optically coupled in cascade, and a first and a second optical differential delay device, represented respectively by blocks 110, 111 , optically coupled to, and interleaved between, the first, second and third splitting device 106, 107 and 108. Exemplarily, the first optical differential delay 110 may be optically coupled between the first and the second splitting device 106, 107 and the second optical differential delay 111 between the second and the third splitting device 107, 108. The device 100 also comprises at least a first input port 101 and at least a first ("bar") and a second ("cross") output port 102, 103 optically coupled to the first input port 101. The device 100 may also comprise a second optional input port 104.

The optical splitting devices 106, 107 or 108 may be any kind of device apt to split an input optical radiation into at least two optical radiations outputting from at least 2 separate output positions of the device. The optical splitting devices 106, 107 or 108 may be for example N x M-port devices, wherein M is at least equal to 2 for all devices and N is at least equal to 2 for devices 107 and 108. For example, the optical splitting device 106, 107 or 108 may be a multi-layer beam splitter or a Fabry-Perot cavity or any wavelength selective mirror (e.g. a Brewster angle window). For example, the optical splitting device 106, 107 or 108 may be or may comprise the optical band splitter/combiner 100 in a nested configuration. In a preferred configuration, the optical splitting device may be a MZI splitter/combiner or an optical coupler, such as for example a MMI coupler or a PLC optical coupler. In a more preferred configuration it may be a single 2-port PLC optical coupler of the kind, preferably, of figure 1. Optical splitting device 106 may also be, for example, an Y-branch coupler. Optical splitting device 106 may have an input port optically connected to the first input port 101 of device 100. Optical splitting device 108 may have a first output port optically connected to the first output port 102 of device 100 and a second output port optically connected to the second output port 103 of device 100.

Independently from the structure or the number of ports, each of the splitting devices 106, 107 and 108 may be viewed, for the purpose of the present invention, as an optical device comprising, with exemplary reference to figure 1 , a first input port 7, a first (bar) output port 8 and a second (cross) output port 6. In this context, optical splitters 107 and 108 also have a second input port 5. At a given wavelength, the power coupling ratio C of a generic splitter is defined according to equation (1). Generally, the power coupling ratio C of a power splitter may be expressed by equation (2). Each of the splitting devices 106, 107 and 108 may have an associated angular expression of the coupling ratio θ_x(λ), x=1,2,3, as shown in figure 4. From (2), coupling coefficients which differ by a multiple of π produce, at a given wavelength, an equivalent effect. Accordingly, the term "net coupling coefficient" will mean a coupling value in the interval from 0 to π equal to the actual coupling coefficient modulo π. The first optical differential delay device 110 and the second optical differential delay device 111 have an associated differential delay, or phase difference, φ_x(λ), x=A, B as shown in figure 4. The first optical differential delay device 110 and the second optical differential delay device 111 may comprise at least a first and a second optical branch disposed in parallel configuration and having different optical-path length, in order to introduce a differential delay, or phase difference, φχ(λ), x=A, B, between the optical radiations propagating through the two optical branches. Generally, the phase difference changes when changing the wavelength of the optical radiations propagating therethrough. For a generic differential optical path delay, typically the longer the wavelength, the smaller the corresponding absolute value of the phase difference: \Φ(^} > \ M if < (3⁾

For the purpose of the present invention, a generic optical differential delay device 110 or 111 may be viewed as an optical device comprising a first input port and a first output port optically connected through the first optical branch and having a second input port and a second output port optically connected through the second branch.

In a preferred configuration, the optical differential delay device 110 or 111 may comprise a pair of planar waveguides having different optical paths. The phase difference can be obtained by adiabatically changing the waveguide shape (e.g. enlarging or narrowing its width) on one arm or by providing an extra length on one arm. By placing the physical change of the waveguide on the opposite arm, the actual phase difference changes the sign.

The phase difference φ may be given by the relation: φ= β₂ L₂ - βi Li wherein L_x, x=1 ,2, is the physical length of the optical path, β_x, x=1,2, is the propagation constant of the optical radiation and the suffixes 1 and 2 refers conventionally to the first branch and the second branch, respectively. As a result, a "positive" actual phase difference will correspond to an optical path of the second branch longer than that of the first branch, while a "negative" actual phase difference means the opposite. Conventionally, the first and second branch of the first optical differential delay device 110 are selected so as to determine an actual value of the phase difference φ_A(λ) which is positive in sign, within the optical bands of interest. The first and second branch of the second optical differential delay device 111 are consequently determined by the optical connection therebetween. More in detail, the optical splitting device 107 has a first input port optically coupled to the first output port of the first differential delay device 110 and a second input port optically coupled to the second output port of the first differential delay device 110. The subsequently cascaded second optical differential delay device 111 has its first input port optically coupled to the first (bar) output port of the preceding splitting device 107 and its second input port optically coupled to the second (cross) output port of the preceding splitting device 107.

The subsequently cascaded third optical splitting device 108 may have its first input port optically coupled to the first output port of the preceding differential delay device 111 and its second input port optically coupled to the second output port of the preceding differential delay device 111. At a given wavelength, phase differences φ which differ by a multiple of 360° (e.g. φ=α+ nx360°, n integer) produce the same effect on the traveling radiation, in the context of the present invention. Accordingly, the term "net phase difference" will mean a phase difference value in the interval from 0° to +360° equal to the actual phase difference modulo 360°. Exemplarily, a (negative) actual phase difference φ = -120°- 2 x 360° will correspond to a net phase difference of 240°. A "zero order" differential delay will mean a differential delay having an absolute value of the actual phase difference smaller than 360° (|ψ| < 360°). A "higher order" differential delay or differential delay of "order n" will mean a differential delay having an absolute value of the actual phase difference greater than or equal to 360° (jφ| > 360°), e.g. φ= α + nx360° or φ= - α - nx360°; α>0 and n>0, n integer.

Optical device 100 of figure 4 is apt to split an input optical radiation P_in into two output optical radiations P_bar and P_cross outputting respectively from bar port 102 and cross port 103 and having a respective power spectrum. Optical device 100 of figure 4 is apt to split/combine a first and a second optical wavelength band in that, considering a large spectrum optical radiation comprising two optical bands as the input optical radiation P_in, when comparing the two power spectra normalized at their maximum intensity, the power spectrum of the cross output radiation P_Cross shows at each wavelength of the first band an optical power greater than the optical power at the same wavelength in the power spectrum of the bar output radiation P_bar- Similarly, the bar output radiation (P_bar) power spectrum shows at each wavelength of the second band an optical power greater than the optical power at the same wavelength in the power spectrum of the cross output radiation Pc_ross- The wavelengths belonging to the first optical band may be shorter than the wavelengths belonging to the second optical band, or vice versa, or the first and second optical band may be spectrally interleaved. Typically it may exists at least one wavelength wherein the two power spectra cross-over. This wavelength is a separation wavelength between the two bands. It is advantageous to define a predetermined level of cross-talk for each of the bar and cross port 102 and 103 to be satisfied by a sub-band of the second optical band and a sub-band of the first optical band, each of them comprised within the respective optical band. Having set a predetermined relative-power level X (dB), which represents a specification on the level of crosstalk at an output port 102 or 103, at each wavelength within the respective sub-band the power level of the power spectrum must be below the predetermined level of cross-talk. Exemplarily, at the cross output port 103, the power at any wavelength within a sub-band of the second optical band (called "stopband") should be X dB below the power of any wavelength within a sub-band of the first optical band (called "passband"). A similar specification holds for the bar port 102. The first input port 101 is preferably apt to receive an optical radiation P_in, for example an optical signal comprising a first wavelength λi and a second wavelength λ₂ comprised into the first and the second optical band, respectively. Typically, the first wavelength λi and second wavelength λ₂ are also comprised into the respective optical sub-band. The first and the second wavelength λi, λ₂ are typically widely spaced, for example the spacing between the first and the second wavelength is greater than about 50 nm or greater than about 100 nm. The sub-bands of the first and the second optical band, for a given level of crosstalk X, are typically wide bands, for example setting a level of cross-talk at -10 dB, they may have a bandwidth greater than or equal to 20 nm, more typically greater than or equal to 40 nm, even more typically greater than or equal to 80 nm. Preferably, the first and the second optical band contain the wavelengths of 1310 nm and 1520 nm, respectively or in inverse order.

In a preferred configuration the first, second and third splitting devices 106, 107 and 108 are apt to perform substantially the same, or equivalent, function. Advantageously, each of the splitting devices 106, 107 or 108 is substantially a full (0-100) splitter at least at a first wavelength λ-ι_op within the first optical band and substantially a null (100-0) splitter at least at a second wavelength λ_2op within the second optical band. For the purpose of the present invention, with reference to a generic optical splitting device, a "full splitter", at a given wavelength λ, will mean a splitter apt to direct the optical power at the wavelength λ, inputting in its first optical port, completely in the cross output port. Similarly, a "null splitter", at a given wavelength λ, will mean a splitter apt to direct the optical power, having wavelength λ, inputting in its first optical port completely in the bar output port. Full and null splitters are also known as "0-100" and "100-0" splitters, respectively. For example, an optical splitter ruled by equation (2) is a full coupler if θ(λ) =π/2+nπ and a null coupler if θ(λ) =nπ, wherein n is an integer.

Preferably, the first, second and third splitting devices 106, 107 and 108 show an identical wavelength dependence of the coupling coefficient θ_x(λ), x=1 ,2,3, within the first and the second optical band. In a more preferred configuration, they are structurally substantially identical, being contemplated that, in general, a difference between two optical devices due to the unavoidable fabrication errors does not depart the optical devices from being substantially identical. Advantageously, the first optical differential delay device 110 has an associated first net phase difference φ_A(mod360°) substantially between about 100° to 140° at least at a wavelength λ_3op within the first optical band and at least at a wavelength λ_4op within the second optical band. The values of the net phase differences at λ_3op and λ_4op typically differ from each other because of the unavoidable wavelength dependence of the phase difference φ. The second differential optical delay device 111 has an associated second net phase difference φ_B(mod360°) substantially between about 220° to 260° at least at a wavelength λ_5op within the first optical band and substantially between about 100° to 140° at least at a wavelength λ_6op within the second optical band. Table 1 and 2 show preferred sets of values of the real phase differences φ_A and φ_B according to the present invention, respectively for the cases _op < λ_op2 and

Table 1

Table 2

It is noted that conventionally the values of the actual phase difference φ_A are positive in sign in the bands of interest, while the values of the actual phase difference φ_B may have a positive or negative sign.

The angular center values of the phase difference ranges shown in tables 1 and 2 represent the ideal values of the phase differences for all the wavelengths within the respective first and the second optical band. Nevertheless, when designing a real optical band splitter/combiner, one must take into account the wavelength dependence of the differential delays, which imposes some constraints to the actual values of the phase differences. Accordingly, when considering the wavelength dependence law of the phase difference, a solution near to the relations of table 1 and 2 are sought through a fine adjustment of the phase differences involved, as will be illustrated below. Similarly, the actual values of the operating wavelengths λ_xop, x=1 to 6, are found in the fine tuning process and may differ from the ideal values at the respective center of the first and the second optical band.

Advantageously, the first optical differential delay device 110 is a zero order differential delay (m=m'=0 in table 1 or 2). This allows to keep the wavelength dependence of the differential delay 110 as low as possible, or equivalent^ to keep the difference between φ_A(λ_3op) and φ_A(λ_4op) as low as possible. Exemplarily, the present invention contemplates embodiments having a difference between φ_A(λ_3op) and φ_A(λ_4op) between about 15° and 40° when the wavelengths λ_3op and λ _op are spaced apart by more than 100 nm. Advantageously, the second optical differential delay device 111 is a higher order differential delay (n>0 in tables 1 or 2). This allows to exploit the wavelength dependence of the differential delay 111 so has to obtain different optical delays φB(λ_5op) and φ_B(λ_6op) at the two optical bands.

The actual choice of indexes n and n' depends on the specific wavelength dependence of the phase difference φ_B(λ) of the differential delay 111. A degree of freedom lays in the choice of the value n'-n. In general, it is advantageous to choose n and n' in order to have the differential delay at the lowest suitable order. An advantage of the low-order solutions is that the wavelength dependence is sufficiently high to make the phase difference φ_B(λ) different at λ_5op and λ_6op and sufficiently low not to destroy the flatness of the pass and stop bands and the stability of the filter. Advantageously, n'=n for the configurations corresponding to first column of Table 1 and second column of Table 2 and n'=n+1 for those corresponding to second column of Tablel and first column of Table 2. Applicant has found that the suitable lowest order choice for differential delay device 111 for the cases, respectively, λ_lop < λ_lop and λ_Xop > λ_lop corresponds to first column of table 1 and second column of table 2 with n'=n, respectively. Table 3 shows these preferred configurations for the respective cases.

Exemplarily, the wavelength dependence of the phase difference φ(λ) may be, in first order, in accordance to the relation:

Φ _,

For band separations Δλ much smaller than the wavelength, the phase difference change in a zero order differential delay can be negligible, due to the moderate value of φ. Conversely, for a sufficiently higher order differential delay (φ sufficiently high), an appreciable difference in the phase differences may occurs. In this way it is possible to realize optical differential delay having suitable wavelength response. Exemplarily, table 4 corresponds to table 1 when inserting the relation: ^6op ^5op φB(λ_6op) " φB(λ₅op)= " φB(λ_5θp) _* ≡ ' φB(λ₅op) Λ (5), λ '5op derived from relation (4). Table 4 is exemplarily restricted to the ideal values of phase differences φ_B(λ_5op) and φ_B(λ_6op) of table 1 corresponding, respectively, to the choice of the net phase difference values of 120° and 240°.

A degree of freedom lays in the choice of the value n'-n. It is clear that choosing n -n for the configuration corresponding to the first column and n'=n+1 for that of the second column, the corresponding value of φ_B(λ_50p) is the lowest possible, and correspondingly the differential delay device 111 is at the lowest suitable order. After the choice of n'-n has been done, it is possible to get an insight of the order n of the differential delay device 111 , as shown exemplarily in table 5 for the preferred choice n -n and n -n+1 , respectively.

In table 5 the expression INT[^*] means the integer part of the argument. From table 5 it is clear that the configuration of the first column is preferred over the second column, in that the order n of the differential delay device 111 is typically lower. Similar considerations for the case λ_lop > λ_2op bring to the conclusion that the second column of table 2 represents a preferred choice for this case.

The choice of the actual values φ_B(λ_5op), φB(λ_6op) and λ_5op, λ_6op is done through a fine adjustment of the values involved in order to find the best filter rejections. Exemplarily, starting from table 5, first column, one may calculate the passbands and stopbands for the overall optical band splitter/combiner corresponding to a choice of the wavelengths λ_5op and λ_6op equal to the center of the respective optical bands. If the passband and stopband are not exactly centered on the sub-bands of interest, one may finely tune the values of λ_5op or λ_6op , thus tuning the value Λ, or finely tune the value at the numerator of the right-hand side of the first equation of the last row of table 5, which comes from the difference between φβ( ₅₀p) and φ_B(λ6_oP). The process stops when an optimal choice of the values φ_B(λ5₀p), φ_B(λ_6op) and λ_5op, λ_6op is found for which the passband and, especially, the stopband are centered over the optical sub-bands of interest.

Optical device 100 may advantageously be on a substrate 120, such as for example a silicon or oxide substrate. In a more preferred configuration, optical device 100 is a PLC optical device, more preferably a PLC optical device comprising a pair of optical waveguides, even more preferably a high density PLC optical device. Advantageously, the refractive index contrast of the waveguides is greater than about 1%, preferably greater than, or equal to, about 2%. Advantageously, the refractive index contrast of the waveguides is lower than, or equal to, about 4.5%, preferably lower than, or equal to, about 3%.

In the following, the principle underlying the present invention, i.e. the principle on which is based the design of a band-flattened structure that is insensitive to moderate changes of the structural parameter, will be illustrated by way of the geometrical representation of the action of splitters and phase delays, as introduced in figures 2 and 3.

Figure 5 shows a geometrical representation of the action of the optical device of figure 4 when an optical radiation P_in having wavelength λ within the first optical band and close but not exactly equal to λ_1op, is fed to the first optical splitting device 106 in its first input port (P_in=Eι). This radiation is mostly, but not completely, directed to the cross output port of the splitter 106. Curve 130 in figure 5 represents the corresponding rotation on the sphere which is close but not exactly equal to π. In figures 5 and 6 only the net angular coupling coefficients and the net phase differences are shown, even thought the present invention contemplates any value of the actual angular coupling coefficient and the actual phase difference which corresponds to the net one. Then the radiation is fed to the first optical differential delay device 110 and undergoes a (net) rotation of about +120°, represented by curve 131. An ideal (net) value of the rotation would be +120°, but the wavelength dependence of the phase difference as well as fine- tuning considerations for the inventive device may deviate the actual value from the ideal value by as much as 40°. The radiation is fed to the second optical splitter 107 and undergoes a further rotation on the sphere which is substantially identical to the rotation caused by the first splitter 106, represented by curve 132. Then the radiation is fed to the second optical differential delay device 111 and undergoes a (net) rotation of about +240°, represented by curve 133. An ideal (net) value of the rotation would be +240°, but the wavelength dependence of the phase difference as well as optimization considerations for the inventive device may deviate the actual value from the ideal value by as much as 40°. Finally the radiation is fed to the optical splitter 108 and undergoes another rotation, represented by curve 134, which is substantially identical to the rotation caused by the first and second splitter 106, 107. From figure 5 it may be explained how the device elements 106, 107, 108, 110 and 111 may possibly cooperate in order to correct the deviation of the splitting ratio of the individual splitters 106, 107 and 108 due to the deviation of the transmitted wavelength λi from the operating wavelength λι_op. Figure 6 shows a geometrical representation of the action of the optical device of figure 4 when an input optical radiation P_in having wavelength λ₂ within the second optical band and close but not exactly equal to λ_2op is fed to the first optical splitting device 106 in its first input port (P_in=E₁). This radiation is mostly, but not completely, directed to the bar output port of the splitter 106. Curve 140 in figure 6 represents the corresponding rotation on the sphere which is close but not exactly equal to 2π. Then the radiation is fed to the first optical differential delay device 110 and undergoes a (net) rotation of about +120°, represented by curve 141. An ideal (net) value of the rotation would be +120°, but the wavelength dependence of the phase difference as well as fine-tuning considerations for the inventive device may deviate the actual value from the ideal value by as much as 40°. The radiation is fed to the second optical splitter 107 and undergoes a further rotation on the sphere which is substantially identical to the rotation caused by the first splitter 106 and represented by curve 142. Then the radiation is fed to the second optical differential delay device 111 and undergoes a (net) rotation of about +120°, represented by curve 143. An ideal (net) value of the rotation would be +120°, but the wavelength dependence of the phase difference as well as optimization considerations for the inventive device may deviate the actual value from the ideal value by as much as 40°. Finally the radiation is fed to the optical splitter 108 and undergoes another rotation, represented by curve 144, which is substantially identical to the rotation caused by the first and second splitter 106, 107. From figure 6 it may be explained how the device elements 106, 107, 108, 110 and 111 may possibly cooperate in order to correct the deviation of the splitting ratio of the individual splitters 106, 107 and 108 due to the deviation of the transmitted wavelength λ₂ from the operating wavelength λ_2op. From the above, it comes that reciprocating the positions of the first and the second differential delay device 110, 111 does not result in any change in the functional behavior of the device.

In use, an optical radiation Pj_n having wavelength λi and λ₂ respectively within a first and a second optical band is fed to the optical device 100 of figure 4 at the input port 101.

The first optical splitting device 106 splits the optical radiation inputting its first input port in such a way that at λ^λi_op it sends the power completely in its respective cross port and at λ₂=λ_2op it sends the power completely in its respective bar port. When λi and λ₂ depart respectively from λ-ι_0p and λ_2op, i.e. for respective splitter rotations slightly different from π and 2π, respectively, Applicant has found that it is possible to correct for this deviation by using two differential phase delays. The first differential phase delay has a net ideal value of 120° both at the first and the second inputting wavelengths and the second differential phase delay has net ideal values of 240° at the first input wavelength λi and of 120° at the second input wavelength λ₂. The result is a flattened response both in the bar port and in the cross port. More particularly, at the bar port 102 the output signal P_bar comprises most of the optical power at the wavelength λ₂ and the optical power at the wavelength λ-ι is kept below the specified crosstalk level in case . Similarly, at the cross port 103 the output signal P_crass comprises most of the optical power at the wavelength λi and the optical power at the wavelength λ₂ is kept below the specified crosstalk level. Exemplarily, the cross talk level may be -10 dB both at the bar and at the cross port. Figure 7 is a schematic planar view of an exemplary PLC optical device 300 in accordance with a preferred embodiment of the optical device 100 of the present invention. The optical device 300 may be a 2-port optical band splitter/combiner for splitting/combining a first and a second optical band comprising exemplarily the wavelengths of 1310 nm and 1550 nm, respectively. Exemplarily, respective sub- bands of first and second optical band are set equal to 1310±50 nm and 1520±40 nm, respectively. A specific level of cross-talk may be predefined for such sub- bands.

Optical device 300 comprises two substantially identical planar optical waveguides 321 , 322 which are advantageously put in close proximity at three different locations over a length L_c in order to obtain three optical couplers 306, 307 and 308. Two optical differential delay devices 310, 311 are interleaved between the coupling regions. In particular, the two optical-path lengths of each of the optical waveguides 321 and 322 in the region between two successive directional couplers 306, 307 or 308 are different from each other in order to build optical differential delays 310 and 311. The first optical differential delay device 310 comprises a first and a second optical arm 310' and 310" and the second optical differential delay device 311 comprises a first and a second optical arm 311 ' and 311 ". Advantageously, a suitable enlargement of the second arm 310" of the first optical differential delay device 310 increases the effective index of the optical mode propagating through it. The corresponding slowing of the radiation passing through the second arm 310" with respect to the radiation passing through the first arm 310' produces an increase of the optical-path length in the second arm 310" and, consequently, a phase difference φ_A. It must be emphasized that the convention introduced above requiring that the phase difference φ_A(λ) is positive within the bands of interest determines that the upper arm 310" of the first optical differential delay 310 is conventionally the second arm. As a consequence, the upper arm 311" of the second optical differential delay 311 is also regarded as the respective second arm, being optically connected to the bar output port of the optical splitter 307, which in turn is the bar port with respect to the input port of the optical splitter 307 connected to the second arm 310" of the first differential delay 310. In FIG. 7, optical waveguides 321 and 322 may be placed on a flat substrate 330. One end of the optical waveguide 321 is a first input port 301 apt to receive an optical signal P_in, and the other end of the optical waveguide 321 is a bar- output port 302 apt to emit a bar optical signal P_bar. Similarly, one end of the optical waveguide 322 is a second input port 304, and the other end of the optical waveguide 322 is a cross-output port 303 apt to emit a cross optical signal P_cross- Exemplarily, the two waveguides 321 , 322 may be buried, ridge or rib waveguides on a substrate material 330 or they may be photonic crystal waveguides on a substrate material. Advantageously, the core to cladding structure of the two waveguides 321 , 322 may be made of a combination of materials such as SiO₂, Ge:SiO₂, BPSG, GBSG, SiON, Si₃N₄, Si, SiGe, Al_xGa.,-_xAs, ln_xGaι-_xAsP, Cd_xZnι-_xTe, GaN or the like or polymeric materials such as polyimides, acrylates, polycarbonates, silicones, benzocyciobutene (BCB), epoxy resins or the like. Figure 8 is a schematic cross-sectional view of an exemplary configuration of waveguide 322 taken along the line A-A in figure 7, wherein the same reference numerals are used where appropriate. The planar optical waveguide 322 is exemplarily an optical waveguide buried into a silica (SiO₂) layer 340 on a silicon substrate 330, such as for example a silicon wafer having a thickness T of 600 μm (figure 8 may not be to scale). The waveguide core-to-cladding index contrast is advantageously chosen equal to about 2.5% at 1550 nm, and it is obtained with a convenient doping of Boron and Phosphorus. Applicant has found that this index contrast is an optimal choice in order to guarantee bending radii of the order of 1.5 mm. An advantage of this solution may be the possibility to achieve smaller devices, i.e. a higher density on a wafer. This is the reason why Applicant has found an optimal choice of index contrast to be higher than commonly used 0.7% index contrast, more preferably higher than about 1%, even more preferably higher than about 2%. On the other hand high wafer density doesn't necessarily mean higher yields, because smaller features and higher index contrast, in general, worsen both fabrication and coupling tolerances. Even thought higher index contrast are feasible, preferably the index contrast is kept below about 4.5%, more preferably below about 3%.

Waveguide 322 has preferably a square core having W x W cross-section width of about 2.6 μm x 2.6 μm, in order to satisfy the requirement of monomodality in the lower end of the low-wavelength band (e.g. at 1310nm-50nm=1260 nm) for the chosen index contrast. Quotes A and B in figure 8 are exemplarily 5 μm and 10 μm, respectively.

The three optical couplers 306, 307 and 308 may advantageously be substantially identical. Figure 9 shows a schematic diagram of an exemplary optical coupler 306. The straight coupling length L_c, the waveguide separations S and s, the radius R and angle θ of the input and output curves are selected, according to known design technique, in order to exemplarily achieve a 0-100 coupling at a first wavelength λ_1op within the first optical band and a 100-0 coupling at a second wavelength λ_2op within the second optical band. In the present example, S=11.92μm, s=2.6μm, R=1700μm, θ=3° and L_c is set equal to 1560.8μm in order to have the optimized values λ-ι_op = 1290 nm and λ_2op=1495 nm. Advantageously, the shortest coupler length that satisfies the splitting condition 0-100 at λι_op and 100-0 at λ_2op is selected as the coupler length L_c. Figure 10 shows schematic a top view of a particular exemplary design of the second arm 310" of the first optical differential delay 310, according to a preferred embodiment of the present invention. Advantageously, the enlargement 350 of figure 10 may have a raised cosine profile and may extend over a length L of about 253 μm and a height H of about 3.6 μm. In this exemplary optimized structure it is sized so that the actual phase difference is +120° at 1290nm. Correspondingly, the actual phase difference at 1500 nm is about +100°. These values of the phase differences have been exemplarily selected by way of a fine-tuning process similar to the one described above aimed to center the exemplary filter on the sub-bands of interest. The simulations were performed with convenient small changes of the differential delays and of the coupler length.

Figure 11 shows a schematic diagram of a particular exemplary design of the optical differential delay 311 , according to a preferred embodiment of the present invention. Optical differential delay 311 is obtained by way of shaping the two optical waveguides 321 and 322 in the form of two arcs having the same angle and different radii in order to achieve a second arm 311" having an extra length with respect to the first arm 311 '. Exemplarily, the angle α and the radii π and r₂ (measured with respect to the waveguide axis) are 9.65°, 1700μm and 1714.52μm, respectively. Accordingly, the difference between the second and the first arm lengths 311 ", 311' is set equal to about 2.44 μm in order to achieve an actual differential phase delay of about +120° + 360° at 1530nm. Correspondingly, the actual phase difference at 1290 nm is about +240° + 360°. The net phase differences at 1530 nm and 1290 nm are respectively 120° and 240°. The wavelength dependence of the phase difference φ(λ) of differential delays 310 and 311 is given, in first order, by expression (4). Table 6 summarizes the design parameters for the embodiment of figure 7: λι_op = 1290 nm < λ_2op=1495 nm φ_A>0, φ_B>0 φ_A (λ₃op) =φA (1290) * +120° φ_A (λ₄op) =φA (1500) * +100° m=m'=0 φ_B(λ_5op)= φ_A (1290) * +240° + 360°

φB(λ₆₀p) «φ_B(λ₅o_P)x[1 - Λ]«0.814φ_B(λ_5op) n' = n =1 Table 6 The fabrication of the exemplary optical device of figure 7 may be done through known processes. Exemplarily, after the deposition of a core layer on a buffer layer of silica on silicon, a resist layer may be spinned on the core layer and subsequently exposed to UV light through a mask. After that, the zones, which have been exposed, may be selectively etched. After that, the patterned layer is advantageously covered, for example with BPSG (Boron Phosphor Standard Glass).

It is advantageous to cascade the optical splitter of the present invention in order to further improve the spectral response, for example to further improve the stopband crosstalk.

Figure 12 shows a schematic diagram of an exemplary device 200 comprising a cascade of optical devices 100 in accordance with the present invention. A first device 100 is cascaded with a second and a third device 100' and 100", all devices being in accordance with the present invention. Advantageously, the second device 100' has its first input port 101' connected to the output bar port 102 of the first device 100. The third device 100" has its first input port 101 " connected to the cross port 103 of the first device 100. An optical signal P_in fed to the first device 100 at its first input port 101 is split into two optical signals outputting from the bar and cross output port 102 and 103, respectively. The signal outputting from bar port 102 is fed to the device 100' and is further split into two optical signals outputting from the two output ports of device 100'. The signal of interest is the signal P_bar outputting from the bar port 102' of device 100', which shows a pass-band comprised within the second optical band and a doubly suppressed stopband comprised within the first optical band. Similarly, the signal outputting from cross port 103 is fed to the device 100" and is further split into two optical signals outputting from the two output ports of device 100". The signal of interest is the signal P_or0sS outputting from the cross port 103" of device 100', which shows a pass-band comprised within the first optical band and a doubly suppressed stopband comprised within the second optical band. In this way it is possible to achieve a spectral response of the composite device 200 which is typically the sum of the spectral response of the single device 100, when both are expressed in logarithmic scale and normalized to the maximum intensity. This is done in order to improve the suppression of the sub-band to be stopped, i.e. to reduce the cross-talk. Figure 13 shows a simulation result of the normalized spectral response of the exemplary optical device of figure 12, wherein all the three optical devices 100, 100' and 100" are according to the embodiment of figure 7 and table 6. In figure 13, the curve denoted P_bar represents the spectral power outputting from bar output port 102', and the curve denoted P_cross represents the spectral power outputting from cross output port 103". Figure 13 shows that the optical device 200 of figure 12 complies with the specification of a -22 dB crosstalk (represented by curve X) for the bar output port 102', wherein the output pass-band of interest is the sub-band from about 1480 nm to about 1560 nm. Specifically, any wavelength within the sub-band from 1260nm to 1360nm in the bar output port 102' is suppressed at a power level at least 22dB below the optical power of any wavelength in the sub-band from 1480 nm to 1560 nm:.

Estimated maximum insertion loss over the whole operating band is less than 1dB. Figure 14 shows a simulation result of the normalized spectral responses of a set of one hundred optical devices 200 of the type shown in figure 12, wherein all the three optical devices 100, 100' and 100" are in accordance to the embodiment of figure 7 and their structural parameters are varied, simultaneously for all the three optical devices 100, 100' and 100", around the values exemplarily given above and corresponding to table 6. The structural parameters taken in consideration for statistical variation are the couplers' waveguide separation (quote s in figure 9), the waveguide cross-section width within the coupling section of the couplers (quote W in figure 8), the length of the straight coupling section of the couplers (quote L_c in figure 9), the length of the first differential delay 310 (quote L in figure 10) and the length of the extra-length of the second differential delay 311 of figure11.

Table 7 in the second column shows the values of the relative standard deviations for the structural parameters used in the simulation. These are found to be a good reproduction of the statistical variations of the structural parameters in the manufacturing process. Table 7 in the third and fourth column shows the calculated corresponding relative standard deviations for the coupling coefficients θ and the phase differences φ of splitters and delays, respectively. The error statistics for both the structural parameters and the corresponding coupling coefficients θ and phase differences φ are assumed random gaussian.

Table 7

From table 7, the overall standard deviation on the coupling coefficients θ is about 2% and the standard deviation on the phase difference of the differential delays φ is about 0.1%. Figure 14 shows the simulation results of a set of one hundred devices 200 whose coupling coefficients θ and phase differences φ are varied according to gaussian distributions centered at values of table 6 and with standard deviations of table 7. Figure 14 shows how device 200 meets the requirements of figure 13 also in the presence of manufacturing errors as large as about the triple of the assumed standard deviations, thus showing strong tolerance to manufacturing deviations. Figure 15 shows an exemplary optical device 400 for use in FTTP networks, which makes use of the present invention. The optical device 400 may be an optical network unit (ONU), i.e. a terminal apparatus of the FTTP network, particularly of the triple play networks, on the customer side. Such a kind of apparatus is also known as "triplexer".

The optical device 400 comprises an optical band splitter/combiner 400a in accordance to the present invention (e.g. as device 100, or device 200 or device 300) for splitting/combining a first and a second optical band. Optical band splitter/combiner 400a comprises a first port 401 , a second port 402 and a third port 403. Exemplarily, the first port 401 may correspond to the first input port (101 of devices 100 and 200 or 301 of device 300), the second port 402 may correspond to the bar output port (102 of device 100, 102' of device 200 or 302 of device 300) and the third port 403 may correspond to the cross output port (103 of device 100, 103" of device 200 or 303 of device 300). Other configurations of the ports are possible, for example the second port 402 may correspond to the cross output port. Optical device 400 also comprises an optical transmitter 404, apt to emit an optical radiation having a first wavelength λ₁ within the first optical band (e.g. near 1310 nm), and optically connected to the third port 403, for example through optical waveguide 410. Optical device 400 also comprises a first optical receiver 406 apt to receive a second optical wavelength λ₂ within the second optical band (e.g. near 1550 nm), and optically connected to the second port 402, for example through optical waveguide 411. The optical device 400 may advantageously comprise an additional optical band splitter 405 for splitting a third optical wavelength λ₃ within the second optical band (e.g. near 1490 nm) from the second wavelength λ₂ and an additional optical receiver 407 apt to receive the third optical wavelength λ₃. Optical band splitter 405 may or may not be in accordance to the present invention.

In use, an optical signal P_up having the first wavelength λ^ emitted by optical transmitter 404 is directed, for example through waveguide 410, into the third port 403 of device 400a. It is then mostly directed to the first optical port 401 of device 400a in order to be fed into an optical transmission line 500 in an up-stream direction. An optical signal P_down having the second wavelength λ₂ and propagating through the optical transmission line 500 in a down-stream direction is fed to the optical port 401 of device 400a. It is then mostly directed to the second optical port 402. Optical signal P_<jown '^s then directed, for example through waveguide 411 , to the optical receiver 406.

Advantageously, the optical signal P_down ^a'^so comprises the third wavelength λ₃. When fed to the optical port 401 of device 400a, it is mostly directed to the second optical port 402. Optical signal P_down is then directed, for example through waveguide 411 , into the additional optical band splitter 405 which splits the radiation into two optical signal Pλ₂ having an optical power mostly at the second wavelength λ₂ and Pλ₃ having an optical power mostly at the third wavelength λ₃. Each of the two optical radiations are subsequently received by the respective optical receiver 406 and 407. In a more preferred configuration, optical unit 400 is based on PLC technology in all or part of its components. Preferably, optical element 400a is a PLC device. In a more preferred configuration, optical element 400a is a high index contrast PLC device, for example having index contrast greater than about 1%, preferably greater than, or equal to, about 2%. Advantageously, the refractive index contrast is lower than, or equal to, about 4.5%, preferably lower than, or equal to, about 3%.

The optical devices (100; 200; 400) in accordance with the present invention find particularly useful applications in optical networks apt to distribute telecommunication services to a plurality of customers. For example, optical networks may be triple play networks or passive optical networks or fiber-to-the- premises networks or, more generally, access networks. The services are delivered using at least a first and a second signal having respectively a first and a second wavelength within respectively a first and a second optical band. Exemplarily, the optical network comprises a central station, a plurality of terminal stations, and a plurality of optical links connecting each terminal station to the central station. Each optical link may comprise cabled fibers and may include passive or active optical devices apt to branch, add, route, amplify, attenuate or switch the optical signals propagating through the link. The terminal station may be of the kind described in figure 15, wherein P_up comprises the first signal and P_d0Wn comprises the second signal.

Claims

1. An optical device (100; 300) for splitting/combining a first and a second optical wavelength band, the device comprising: - a first, a second and a third optical splitting device (106, 107, 108; 306, 307, 308) optically coupled in cascade; and a first differential optical delay device (110, 310) optically interposed between the first and second optical splitting device (106, 107; 306, 307) and a second differential optical delay device (111 ; 311 ) optically interposed between the second and third optical splitting device (107, 108; 307, 308); characterized in that: each of the first, second and third optical splitting device (106, 107, 108; 306, 307, 308) is substantially a full splitter at least at a respective first wavelength within the first optical band and a null splitter at least at a respective second wavelength within the second optical band; the first differential optical delay device (110; 310) has an associated first net phase difference (φ_A) substantially between about 100° and 140° at least at a third wavelength within the first optical band and substantially between about 100° and 140° at least at a fourth wavelength within the second optical band; and the second differential optical delay device (111 ; 311) has an associated second net phase difference (φ_B) of between about 220° and 260° at least at a fifth wavelength within the first optical band and between about 100° and 140° at least at a sixth wavelength within the second optical band.

2. The optical device of claim 1 , wherein the first differential optical delay device (110; 310) is a zero order differential optical delay device in the first and second optical band.

3. The optical device of claim 1 or 2, wherein the second differential optical delay device (111 ; 311 ) is a higher order differential optical delay device in the first and second optical band.

4. The optical device of claim 3, wherein the second differential optical delay device (111 ; 311) has the same order in the first and in the second optical band.

5. The optical device of claim 4, wherein the second differential optical delay device (111 ; 311) is a first order differential optical delay device.

6. The optical device of any of the preceding claims, wherein the optical device is a PLC optical device (300).

7. The optical device of claim 6, wherein the optical device comprises a pair of optical waveguides (321, 322) forming the first, second and third optical splitting device (306, 307, 308) and the first and second differential optical delay device (310, 311 ).

8. The optical device of claim 7, wherein the optical waveguides (321 , 322) have an index contrast higher than about 1%.

9. The optical device of claim 7, wherein the optical waveguides (321 , 322) have an index contrast lower than about 4.5%.

10. An optical device (200) for splitting/combining a first and a second optical wavelength band, comprising at least two optical devices according to any of claims 1 to 9 optically connected in cascade.

11. An optical network unit (400), comprising: - an optical device (400a) for splitting/combining a first and a second optical wavelength band according to any of the preceding claims; - an optical receiver (406) optically connected to the optical device (400a) and apt to receive a first signal within one of the first and second optical band; and - an optical transmitter (404) optically connected to the optical device

(400a) and apt to transmit a second signal within the other of said first and second optical band.

12. The optical network unit (400) of claim 11 , wherein the optical device (400a) for splitting/combining a first and a second optical wavelength band comprises an input port (401) apt to be optically connected to an optical transmission line (500) apt to propagate said first and second signal.

13. An optical network suitable to operate at least at a first and at a second optical transmission wavelength respectively within a first and a second optical wavelength band, the network comprising at least an optical device according to any of the preceding claims for splitting/combining said first and second optical wavelength band.