The present invention relates to optical spectrum analysis of an optical signal.
Optical spectrum analysis serves to analyze an unknown spectrum to determine the power of certain wavelengths of the spectrum. For such an analysis it is normally used an optical spectrum analyzer (OSA) which typically comprises a pivotable grating. The grating serves to disperse the light under measurement and so to select a certain wavelength of the spectrum. By pivoting the grating the spectrum can be analyzed. To allocate a certain wavelength to a certain pivot angle it is necessary to calibrate the OSA in advance. For such a calibration sharp laser lines, gas absorption lines or reflection lines of fiber Bragg gratings or Fabry-Perot filters with a known wavelength are usually used.
The resolution bandwidth (RBW) of such an OSA is substantially given by the formula RBW=λ/(n N), with λ being the wavelength, N being the number of grating passes, and n being the number of illuminated grating lines. In known OSAS, such as the 8614×A of Agilent Technologies Inc., a Delaware Corporation, Palo Alto, Calif. 94303, USA, the RBW can be around 40 pm. Generally, there are two principle ways to enhance the RBW of such an OSA. First, to increase the number of the illuminated lines, i.e., Increase the size of the grating, or second, to increase the number of grating passes, i.e., increase complexity.
It is an object of the invention to provide improved optical spectrum analysis of an optical signal.
The object is solved by the independent claims.
The present invention comprises the perception that a prior art OSA has limited wavelength accuracy. In known OSAs the wavelength accuracy is about 10-20 pm. Furthermore, since the calibration result of the OSA is dependent on ambience parameters, e.g. the temperature in the ambience of the OSA, the accuracy of the OSA results are worsening rapidly with time because of a change of such parameters with time.
Moreover, the afore mentioned accuracy of prior art OSAs is only present in the vicinity of about 10-20 nm of a wavelength used for calibration. In greater distances from the calibration wavelength the accuracy is also worsening and can be as low as 100-200 pm.
Since embodiments of OSAs of the present invention use a determination of the wavelength of the filtered light beam, e.g. the output of a diffractive or dispersive element, e.g. a grating, the afore mentioned problems of the prior art OSAs do not occur.
It is therefore an advantage of such embodiments that the accuracy of the finally determined wavelength is about significant higher than of a prior art OSA, i.e., the wavelength accuracy of embodiments of the present invention can be as high as 1-2 pm.
Additionally, the wavelength accuracy of inventive embodiments is independent of any time dependent behavior of the used dispersive element, but is only dependent on the accuracy of the calibration of the wavelength detection unit used for the determination of the wavelength of the output of the dispersive element. This Is because according to embodiments of the present invention the wavelength determination of the output is not restricted to be done in advance of an OSA measurement but can be done parallel to or simultaneously with the action of the dispersive element. Such embodiments therefore provide an inherently and continuously calibrated measurement result due to the Integrated determination of the wavelength of the output of the dispersive element.
Since wavelength detection units can be calibrated with a higher accuracy than known OSAs, embodiments of the present invention enable the afore mentioned strong enhancement of the possible wavelength accuracy down to 1-2 pm.
In a preferred embodiment of the present invention a diffractive grating is used as a dispersive element.
It is preferred to evaluate the wavelength of the output of the dispersive element by determining the center of gravity wavelength or average wavelength of the output. This is advantageous because the normal spectrum to be analyzed is not like a sharp signal of a laser but will show a certain bandwidth. Therefore, the output of the dispersive element will show a certain bandwidth, also. Therefore, it is preferred to determine the average wavelength of the output.
Preferably, to implement the determination of the average wavelength λaverage of the output the output O(λ) is folded with a transfer function T(λ) according to the following formula: λaverage=∫(T(λ) O(λ)) dλ.
The transfer function T(λ) is preferably linear or nearly linear to enhance the accuracy of the average wavelength λaverage.
In a preferred embodiment of the present invention the transfer function T(λ) is realized by using at least two cyclical or periodic functions which are phase shifted with respect to each other. Using such periodic functions is advantageous with respect to a simple linear filter because the wavelength resolution of such functions is higher. However, the use of a simple linear filter is also possible.
To realize such a periodic transfer function an embodiment of the present invention uses at least two etalons which have a slightly different thickness with respect to each other in the direction of the output beam of the dispersive element. Each etalon generates then a periodic output which due to the differing thickness are shifted with respect to each other. Preferably, such output power or transmission power of each etalon is then normalized to the whole detected power from each etalon. On the basis of such normalized transmission power it is then possible within the free spectral range (FSR) of each etalon to precisely determine the wavelength of the output of the dispersive element.
With other words: the etalons and respective power detectors serve as a wavelength detection unit to determine the accurate wavelength of the output of the dispersive element.
An example of such etalons is given in applicants EP 0 875 743 B1 the disclosure of which is fully incorporated herein by reference. However, instead of etalons any other birefringent element, such as retardation plates or wave plates or any other optical element or system which has a periodic or nearly periodic transfer function with respect to the wavelength can be used for the purposes of the present invention. Also, it is possible to use any other known wavelength detection unit, e.g. a Michelson interferometer.
When using such periodic transfer functions it is preferred to select among the periodic peaks the peak within the correct FSR by using a selective element. The correct FSR is the FSR corresponding to the wavelength of the filtered light beam, the FSR, which comprises the wavelength of the filtered light beam. Such a selection roughly indicates the correct wavelength and therefore enables to choose the correct FSR comprising the desired peak. Such a selective element can be an edge filter or any other means indicating a rough estimation of the wavelength.
However, for the present invention it is preferred to use the preselection of the dispersive element to select the correct FSR comprising the desired peak.
Moreover, it is preferred to use a spectral bandwidth of the output signal of the dispersive element, which is, e.g. at least 10 times, smaller than the periodicity of the periodic transfer function to cause interference by the birefringent element.
It is also preferred to use a bandwidth of the output signal of the dispersive element, which is, e.g. at least 20-40 times, smaller than the FSR of the periodic transfer function to be able to have at least one peak within the FSR.
Other preferred embodiments are shown by the dependent claims.
The invention can be partly embodied or supported by one or more suitable software programs, which can be stored on or otherwise provided by any kind of data carrier, and which might be executed in or by any suitable data processing unit. Software programs or routines are preferably applied to the realization of the inventive method.
Other objects and many of the attendant advantages of the present invention will be readily appreciated and become better understood by reference to the following detailed description when considering in connection with the accompanied drawings. The components in the drawings are not necessarily to scale, emphasis instead being placed upon clearly Illustrating the principles of the present invention. Features that are substantially or functionally equal or similar will be referred to with the same reference sign(s).
Referring now in greater detail to the drawings,
An unknown light beam 4 under measurement is coming from a source 3. The light beam 4 under measurement is provided to a lens 6 to collimate the original light beam 4 into a collimated light beam 8. The collimated light beam 8 is provided to a selective element 9 for selecting a certain part of the optical spectrum of the light beam 8 to provide a filtered light beam 12. The selective element 9 comprises as a dispersing element a diffractive grating 10 acting as a variable filter for providing a variable filtering of the collimated light beam 8 for selecting the certain part of the optical spectrum of the collimated light beam 8.
To provide such a filtering the diffractive grating 10 is diffracting different parts of the optical spectrum of the collimated light beam 8 at different angles to provide a spatially dispersed light beam and to select a certain spatial part of the spatially dispersed light beam to finally provide the filtered light beam 12.
The part of the optical spectrum of collimated light beam 8 that will be selected by the grating 10 can be chosen by turning the grating 10 according to arrow 11 of
After such a selection the filtered light beam 12 Is provided to a lens 14 to focus the filtered light beam 12 into a filtered and focused light beam 16. Lens 14 focuses the filtered and focused light beam 16 on a wavelength detection unit 18.
By adding further optical elements the grating 10 can be hit multiple times by the input light beam 4 or the filtered light beam 12 to increase dispersion and thus to improve resolution of the instrument. These more complicated setups are functional equivalent to the simplified structure shown in
The function of the wavelength detection unit 18 is as follows:
To evaluate the wavelength of the filtered and focused light beam 16 which is the output of the grating 10 the center of gravity wavelength or average wavelength of this output is determined. This is done because normally a spectrum to be analyzed will not be like a sharp signal but will show a certain bandwidth. Therefore, the output of the grating 10 will show a certain bandwidth, also.
To implement the determination of the average wavelength λaverage of an output O(λ) of the grating 10 the output O(λ) Is folded with a transfer function T(λ) according to the following formula: λaverage=∫(T(λ) O(λ)) dλ.
The transfer function T(λ) Is nearly linear to enhance the accuracy of the average wavelength λaverage. The transfer function T(λ) is realized by using two periodic functions which are phase shifted with respect to each other. Alternatively, a simple linear filter can be used. To realize such a periodic transfer function T(λ) the wavelength detection unit 18 of the embodiment of
To know the relative transmission of each etalon 1 and 2 as shown in
Alternatively, instead of etalons any other optical elements with two beam or multiple beam interference or any birefringent elements, such as retardation plates or wave plates can be used. Also, it is possible to use any other known wavelength detection unit, e.g. a Michelson interferometer.
Since the etalons provide a periodic transfer function it is necessary in the shown embodiment of
In the embodiment of
The wavelength detection unit 18 provides an electrical signal to the evaluation unit 20 which electrical signal is corresponding to signals of the two etalons I and 2 according to
To cause interference by the etalons 1 and 2 the filtered and focused light beam 16 has a bandwidth, which is, e.g. ten times, smaller than a periodicity of the periodic transfer function of
In the middle of the graph there is shown a third curve 72. This third curve 72 displays a reconstructed or corrected power amplitude of the detected filtered and focused light beam 16 over the relative wavelength. Such a correction can be done if there are some values In the uppermost optical spectrum of the
The graph of
To provide such a second correction evaluation unit 20 evaluates the density of the measured values on the x-axis, i.e., the axis of the relative wavelength. With such a measure it is possible to enhance resolution and to see two peaks In the therefore twice corrected resulting optical spectrum 92 of filtered and focused light beam 16 according to
Therefore, the present invention comprises also measures to provide a second correction of the measured spectrum of filtered and focused light beam 16 so that peaks of the incoming light beam 4 with a distance of as low as 0.5 RBW can be separately displayed in the detected spectrum of filtered and focused light beam 16. This means that the present invention can generate an accuracy of the detected wavelength of as high as 1-2 pm.
The wavelength detection unit is more sensitive to adjacent peaks in the light spectrum under measurement compared to the power detection unit (i.e. the power of adjacent peaks are not detected as different peaks but as one single peak, if the wavelength distance falls within the range of the measurement aperture). In a further embodiment, the resolution of the measurement is enhanced without reducing the aperture. Therefore, the information of the wavelength measurement is used to improve the power measurement resolution. The two signals derived from both the power and the wavelength measurement, or any signals derived there from can be used as Input values for a direct or iterative deconvolution procedure for correcting an impact of a filter shape of the variable filter on the power amplitude as a function of the wavelength of the filtered light beam 12.
However, as Input values for such a deconvolution procedure not necessarily power and wavelength information must be used, but any at least two input signals, derived from applying appropriate different periodic spectral weighting functions to optical signals and subsequent detecting these two signals can be used.
Number | Date | Country | Kind |
---|---|---|---|
PCT/EP04/52401 | Oct 2004 | EP | regional |