Wavelength division multiplexing

Details Wavelength division multiplexing Wavelength division multiplexing

2000-08-04

2002-01-15

Ngo, Hung N. (Department: 2874)

Optical waveguides

With optical coupler

Input/output coupler

C385S046000, C385S024000

Reexamination Certificate

active

06339664

ABSTRACT:

The invention lies in the field of wavelength division multiplexing and relates to the design of arrayed waveguide gratings.
For networks employing wavelength division multiplexing (WDM), for example passive optical networks (PONs), there is a requirement for components for use as wavelength selective multiplexers, demultiplexers and routers. Examples of such components are fibre Fabry-Perot filters, in-line fibre Bragg gratings, free-space diffraction gratings, cascaded filters and arrayed waveguide gratings, the last-mentioned of which components being the subject of the present invention.
An arrayed waveguide grating (AWG) is often referred to as a waveguide grating router (WGR), and sometimes as a phased array (PHASAR) or phased array waveguide grating (PAWG).
As described in U.S. Pat. No. 5,002,350, an AWG can be considered to be made up of two star couplers, one on the input side and one on the output side of the AWG, which are interconnected by an array of M waveguiding channels, in sequence m=1 to M with the channels having gradually increasing optical path lengths such that the optical path length of the mth channel is greater than that of the (m−1)th channel by a fixed increment &Dgr;l.
AWGs have been manufactured from doped silica films deposited on silicon substrates and using indium-phosphide-based technology.
Some AWGs have an optical path length increment &Dgr;l between adjacent channels, or grating arms, which cannot be changed in use, this kind of AWG being referred to as passive in the following. An example of a passive AWG is to be found in U.S. Pat. No. 5,002,350.
U.S. Pat. No. 5,515,460 discloses another kind of AWG which allows the size of the optical path length increment &Dgr;l to be varied in use, this kind of AWG being referred to as active in the following. In the design of the AWG of U.S. Pat. No. 5,515,460, the optical path lengths are varied through the application of an electrical current signal I, i.e. &Dgr;l=f(I), so that the device can be tuned by varying the current of the applied control signal.
In standard AWGs, passive and active, the shape of the passband of the channels is approximately sinc squared, being a multiple convolution of the input and output waveguide modes with the echelon transfer function produced by the waveguide array. In consequence, the wavelength shift tolerance is minimal. That is, if the input wavelength varies unintentionally by a small amount, this is likely to cause an unintended change in the output port to which the input signal is routed, or at least a loss in the signal power coupled into the intended output port.
A number of ways of broadening the passband of an AWG to overcome this problem are known from the following documents:
(1) M. R. Amersfoort et al, Electronics Letters, volume 23, pages 449 to 451 (1996), disclose a device having multimode interference couplers.
(2) Y. P. Ho et al, IEEE Photonic. Tech. Lett. volume 9, pages 342 to 344 (1997), disclose a device having multiple Rowland circles on the output side.
(3) A. Rigny et al, Proceedings 23rd ECOC Edinburgh UK, pages 79 to 82 (September 1997) in IEE Conference Publication No. 448, disclose a device having two interleaved sets of channels, each set having a different optical path difference increment.
(4) U.S. Pat. No. 5,412,744 (Dragone) discloses a double waveguide input in the form of a Y-branch.
(5) D. Trouchet et al, OFC '97 Technical Digest, pages 302 and 303 discloses a device in which the input and output star couplers have two focal points.
According to first aspects of the invention, exemplified by claims 1 to 29 of the attached claims, there are provided arrayed waveguide gratings in which the channels have a non-linear optical path difference progression.
It will thus be appreciated that the present design is fundamentally different from that of conventional AWGs in which the optical path lengths of the channels increase in equal steps with channel number, thus to provide a linear phase profile considered to be an essential technical requirement in conventional AWGs. By contrast, in the present design, the optical path length increment between channels is a function of channel number, i.e. the optical path lengths of the channels progress in non-equal steps with channel number, thus to provide a non-linear phase profile. The non-equal optical path length increment between channels can be thought of as comprising a component of equal path length increments, as in a conventional AWG, in combination with an additional component of path length increments which, in the preferred embodiments, is a super-linear polynomial function. It is this additional component which is the source of the non-linear attributes of the phase profile.
It will be understood that the increment need not be positive, but may also be negative, i.e. a decrement and the use of the word increment when describing the claimed invention and its embodiments should be understood as having this meaning.
In some embodiments the additional component of path length increments follows a parabolic function and thus results in the non-linear phase profile being a parabolic profile, i.e. of order 2.
In other embodiments the non-linear phase profile is super- or sub-parabolic. In one specific example, it is found that a super-parabolic, non-linear profile of order 2.1 provides optimum performance.
Generally, the non-linear profile may contain higher even, and/or odd, order expansion terms, since in general a function can be expanded in terms of even, and/or, odd powers.
It is through the departure from the linear phase profile of conventional AWGs that passband broadening can be achieved without requiring any additional elements and without adding complexity to the waveguiding structure and thus the fabrication.
Optical path length can be varied by varying the geometrical channel length and/or by varying the refractive index along the channels.
According to second aspects of the invention, exemplified by claim 30 to 41 of the attached claims, there are provided arrayed waveguide gratings in which the passband is tailored by amplitude modulation of the channel transmission, for example by channel-dependent attenuation or gain within the channels, or via incorporation of a channel-dependence of the coupling coefficients for coupling into or out of the channels.


REFERENCES:
patent: 5002350 (1991-03-01), Dragone
patent: 5351317 (1994-09-01), Weber
patent: 5367586 (1994-11-01), Glance et al.
patent: 5412744 (1995-05-01), Dragone
patent: 5515460 (1996-05-01), Stone
patent: 5745612 (1998-04-01), Wang et al.
patent: 5862279 (1999-01-01), Amersfoort et al.
patent: A-762163 (1996-12-01), None
patent: A-5-323246 (1993-12-01), None
patent: WO 95 22070 (1995-08-01), None
patent: WO 95/22070 (1995-08-01), None
patent: WO 99 45420 (1999-09-01), None
Farjady F.,Parker MC, Walker SD: “Non-linear phase apodisation techniques for arrayed-waveguide grating passband control” IEEE Colloquium on Multiwavelength optical Networks: Devices, Systems, 17thJun. 1998, pp. 9/1-9/5, XP000783512.
M R Amersfoort et al, Electronics Letters, vol. 32, pp. 449 to 451 (1996), “Passband broadening of integrated arrayed waveguide filters using multimode interference couplers”.
Y P Ho et al, IEEE Photonic. Tech. Lett. vol. 9, pp. 342 to 344 (1997), disclose a device having multiple Rowland circles on the output side.
A Rigny et al, Proceedings 23rdECOC Edinburgh UK, pp. 79 to 82 (Sep. 1997) in IEE Conference Publication No. 448, disclose a device having two interleaved sets of channels, each set having a different optical path difference increment.
D Trouchet et al, OFC '97 Technical Digest, pp. 302, 303.

Affiliated with

Also associated with

Rating

Wavelength division multiplexing does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Wavelength division multiplexing, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Wavelength division multiplexing will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2821070