This invention relates to the field of optical filters and, more specifically, to fast tunable optical filters.
High-speed data communications systems need to support the aggregate bandwidth requirements of current and future applications such as telecommunication technology, supercomputer interconnection, high-quality video conferencing and multimedia traffic. There is a general consensus that these bandwidth requirements can most easily be attained by using optical transmission technologies. Dense optical wavelength division multiplexing (WDM) appears to be the hardware backbone for such networks. Dense optical WDM is a method of multiplexing a large number of optical data channels on a wavelength basis (e.g., each divided wavelength region is regarded as a different channel and is routed and manipulated separately from all other divided wavelength regions).
Dense WDM requires advanced optoelectronic components and subsystems capable of handling the extremely high aggregate bit rates and traffic levels demanded by modern optical data communications systems. One very critical component needed for the implementation of WDM packet-switched systems is a tunable filter. A tunable filter is used to extract one or a subset of optical channels from the multitude of optical channels entering the filter. The subset of extracted channels can be varied in time by dynamically modifying the filter configuration
Some common optical filters are based on classical interferometers, and include Fabry-Perot and Bragg filters. Such filters are tuned by mechanically or thermally moving the resonating structure, and the tuning speed is therefore comparatively slow, typically of the order of milliseconds.
Another type of tunable filter is based on the acousto-optic effect. Such components depend on the interaction between an acoustic wave generated in the device, and the optical signal input into the filter. The tuning is achieved by altering the frequency of the acoustic wave, which can be accomplished by altering the frequency of the electronic signal used to generate the acoustic wave. These filters are, however, polarization dependent, which may require correction and cause various other practical problems.
The present invention advantageously provides a fast tunable optical filter with unique selection means, a reduced form factor (size), and with easier and more efficient operability. The fast tunable optical filter of the present invention is capable of being monolithically integrated on, for example, silica or semiconductive wafers implementing current techniques used for planar lightwave circuit construction.
In one embodiment of the present invention an apparatus includes a waveguide grating router (WGR) for receiving up to P input optical signals and responsively providing up to Q outputs, where P and Q are integers greater than or equal to one, at least one input selecting device for selectively coupling up to P optical signals to the WGR as input signals, and at least one output selecting device for selectively inhibiting up to Q outputs of said WGR. The number of input ports, P, and the number of output ports, Q, of the WGR are determined such that the total number of ports, (P+Q), of said WGR is substantially minimized.
The teachings of the present invention can be readily understood by considering the following detailed description in conjunction with the accompanying drawings, in which:
To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures.
The present invention advantageously provides a fast tunable optical filter with unique selection means that can be realized in integrated form with a reduced number of controls. Although the present invention will be described within the context of a fast tunable optical filter comprising specific components, it will be appreciated by those skilled in the art that alternate components performing substantially similar functions can be implemented within the fast tunable optical filter of the present invention in accordance with the present invention.
The purpose of tunable optical filters is to extract one or a subset of optical channels from a multitude of optical channels entering the filter. Additionally, the subset of optical channels extracted can be varied in time by dynamically modifying the filter configuration.
Although in
Briefly stated, an optical signal comprising at least one optical channel comprising light energy at respective spectral regions enters the fast tunable optical filter 100 via an input waveguide and is divided by the power divider 110. Subsequently, the input shutters 120 either block or allow transmission of the power divided optical signal to the WGR 130. The passed optical signals are coupled to the WGR 130 through respective input ports p (1≦p≦P) and are switched by the WGR 130 and routed thereby to the output ports q (1≦q≦Q) of the WGR 130. The output shutters 140 either block or transmit the switched optical channels from the WGR 130 before reaching the reflectors 150. The reflectors 150 reflect the selected output channels back to the input of the fast tunable optical filter 100. The reflected optical channels can be separated from the incident optical signals by, for example, an optical circulator (not shown).
More specifically, in the fast tunable optical filter 100 of
A WGR is a planar lightwave circuit comprising P input and Q output ports separated by two slab regions and connected by M waveguides with varying lengths. The inventors have determined that WGRs represent excellent solutions for providing large optical cross-connects. They are fully passive elements and can provide strictly non-blocking connections for a set of N optical channels. In the fast tunable optical filter 100 of
In accordance with the present invention, a combination of a minimum number of input ports and output ports of the WGR 130 provide a route for the optical channels of the input optical signals through the WGR 130. The input ports and the output ports of the WGR 130 are selected to reduce the size of the WGR 130 and subsequently, the fast tunable optical filter 100, and to reduce the power loss of an optical signal through the WGR 130. An inventive method for determining the numbers of input ports and output ports of the WGR 130 and their geometric location on the WGR 130 will be described in detail below.
Referring back to
In the fast tunable optical filter 100 of
The inventors have developed an inventive method for determining the number of inputs, P, and the number of outputs, Q, for the WGR 130 that minimizes the number of controllers (i.e., input shutters 120 and output shutters 140) necessary to perform the channel selecting function of the fast tunable optical filter 100. Solving for the solution that minimizes the sum P+Q is advantageous because, as such, the number of input shutters 120 and output shutters 140 to be controlled at the input and output ports of the WGR 130 is minimized. When the number of controllers required in the WGR 130 results in a reduction of both the form factor (size) of the WGR 130 and the size of the fast tunable optical filter 100, and ultimately increases the yield and manufacturability of the device resulting in a cost reduction.
The routing properties of a WGR with P input ports and Q output ports allow for P×Q possible connections and, because in the present invention each connection is associated with a unique optical channel (wavelength), P×Q must be larger or equal to N if N optical channels have to be supported by the WGR. As such, N≦P×Q, where N is the number of optical channels to be supported by a WGR, P is the number of input ports of the WGR, and Q is the number of output ports of the WGR.
For any real number Z=W×Y, the minimum sum W+Y is obtained when W=Y=√{square root over (Z)} (this follows by computing the derivative
such that W+Y=2√{square root over (Z)}. As such, for any other combination of W, Y the sum W+Y will not be minimal. Furthermore, for any Z*>Z, the minimum sum W*+Y*=2√{square root over (Z)}>2√{square root over (Z)}. In the integer domain, for any integer B>A (where integer A=C2, C being an integer), the minimum of the sum b1+b2 (with b1×b2=B) is larger than the minimum of the sum a1+a2 (with a1×a2=A). As such, the minimum sum a1+a2=2C. Also, for any B>A, the integer minimum sum b1+b2≧2√{square root over (B)}>2C.
Using the above concepts for the case of evenly spaced optical channels λk, the minimum total number of shutters P+Q is determined from the solutions of the equation (1) as follows:
N≦P×Q≦[CEIL(√{square root over (N)})]2 (1)
where the ceiling function, CEIL, denotes the next higher integer for a non-integer argument and CEIL(x)=x for an integer x, P is the number of input ports of the WGR, and Q is the number of output ports of the WGR. The right hand side of equation (1) above is the next higher integer that can be found after N and that can be expressed as C2 (C being an integer). Therefore if no product P×Q (where N≦P×Q<[CEIL(√{square root over (N)})]2=C2) has P+Q<2C, then the minimum sum will be P+Q=2C and P=Q=C for the reasons stated above. Equation (1) restricts to a finite number of pairs (P,Q) among which the solution for the minimum sum P+Q has to be found. This greatly speeds up the search for the optimum solutions for the number of input ports, P, of the WGR and the number of output ports, Q, of the WGR. Once the solution pairs (P,Q) have been determined for the WGR, a power loss associated with a power splitter (such as the power splitter 110 of
For unevenly spaced optical channels, the determination of the minimum number of input ports, P, of the WGR and the number of output ports, Q, of the WGR is more complex because the number of ports of the WGR strongly depends on how the channels are individually positioned across the spectrum. The minimum total number of input ports and output ports of the WGR, P+Q, is determined from the equations as follow:
N*≦P×Q≦[CEIL(√{square root over (N*)})]2 (2)
where
As in equation (1) above, in equation (2) the ceiling function, CEIL, denotes the next higher integer for a non-integer argument and CEIL(x)=x for an integer x, P is the number of input ports of the WGR, and Q is the number of output ports of the WGR.
In equation (3) above, the Max and Min functions denote the maximum and minimum wavelength values, respectively, of the optical channels, λk (k=1 . . . N), and LCF denotes the largest common factor among the wavelength differences (λj−λl) between the N optical channels for j, l=1 . . . N. The values of λk (k=1 . . . N) can be adjusted to ensure that the LCF function does not become too small and consequently N* too large. If N* is not an integer, N* will have to be rounded to the next higher integer. It should be noted that if the wavelengths λk (k=1 . . . N) are equally spaced then N=N*. As with the case of evenly spaced channels, solving for the solution that minimizes the sum P+Q is necessary to minimize the number of input and output ports of the WGR 130 to be controlled. As in the case of evenly spaced channels, once the solution pairs (P,Q) have been determined, the power loss due to a power splitter is minimized by selecting P≦Q.
It should be noted though, that although equations (2) and (3) above always generate a solution for the sum P+Q, the solution does not always depict a minimum for the number of input ports, P, and the number of output ports, Q. An alternative method to obtain a smaller sum of P+Q consists of splitting the total number of optical channels (wavelength sequence) into subsequences each one with a constant channel separation. The method described above and applied for equally spaced channels is then applied to each subsequence separately. For both cases, for evenly and unevenly spaced channels, the minimum sum of the number of input ports, P, and the number of output ports, Q, follows the relation P+Q≦N+1.
Once the number of inputs, P, and outputs, Q, for a WGR are determined, a geometric location for the input ports pj (j=1 . . . P), and the output ports qh (h=1 . . . Q) of the WGR must be determined so as to reduce a power loss and enhance the loss uniformity associated with the WGR.
By locating the input ports p1-p4 and the output ports q1-q8 as close to the center of the Brillouin zone as possible, the loss uniformity among the ports is optimized. As such, a power loss associated with the function of the WGR 130 in the fast tunable optical filter 100 is reduced. The imaging properties of the WGR 130 behave in a manner such that, the more closely spaced the input ports, the larger the spacing between the output ports needs to be when all N channels are correctly accomodated.
Therefore, by increasing the distance between the input ports, the output ports will move closer. This procedure is iterated until the total span covered by the input ports pj approximately equals the total span covered by the output ports qh. In such a manner, the most central part of the Brillouin zone is used allowing for optimized transmission loss and loss uniformity across the WGR 130. The improved performance of a FRD (illustratively a WGR) in accordance with the present invention is further advantageous in that it decreases the operating losses of a fast tunable optical filter. As such, the fabrication tolerances can be relaxed resulting in a manufacturing yield increase and cost reductions.
The fast tunable optical filter 200 of
In the fast tunable optical filter 300 of
The number of inputs and outputs to the WGR 320 of the fast tunable optical filter 300 of
In the fast tunable optical filter 300 of
While the forgoing is directed to various embodiments of the present invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof. As such, the appropriate scope of the invention is to be determined according to the claims, which follow.
Number | Name | Date | Kind |
---|---|---|---|
5136671 | Dragone | Aug 1992 | A |
5488500 | Glance | Jan 1996 | A |
5543010 | Keng | Aug 1996 | A |
5675592 | Dragone et al. | Oct 1997 | A |
5701371 | Ishida | Dec 1997 | A |
6222964 | Sadot et al. | Apr 2001 | B1 |
6243402 | Doerr | Jun 2001 | B1 |
6359912 | Doerr et al. | Mar 2002 | B1 |
6381383 | Bernasconi et al. | Apr 2002 | B1 |
6411756 | Sadot et al. | Jun 2002 | B2 |
6584244 | Hung | Jun 2003 | B2 |
Number | Date | Country | |
---|---|---|---|
20040096154 A1 | May 2004 | US |