Method and system for determining in-band optical noise
A technology with in-bandwidth and noise components, applied in transmission systems, electromagnetic wave transmission systems, transmission monitoring/testing/fault measurement systems, etc., can solve problems such as increasing modulation rate and difficulty in reliable measurement
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0072] will now refer to Figure 7 , describing an example implementation of the processing performed by the spectral processor 18 in order to discriminate between the signal component S and the noise component N.
[0073] In the following, the spectral traces are analyzed and processed in the spectral neighborhood of the DWDM optical channel or each signal peak included into the input optical signal for which the in-band noise level is to be determined. Signal peaks are identified using standard known implementation techniques to determine their respective center wavelengths (λ p ). The in-band noise level and OSNR on each signal peak can be estimated using the methods described herein without requiring the acquisition of other further spectral traces.
[0074] In step 502, different polarization analysis conditions are used as described above, and in this case generally Figure 6 The polarizer 34 generates two samples p from the input optical signal p A and p B . In st...
Embodiment 2
[0138] now refer to Figure 10 and Figure 11 , describing an example implementation of a method according to the DRBD method.
[0139] In this example, the numerical integration function h 1 (λ) and h 2 (λ) are respectively with width RBW 1 and RBW 2 The rectangular convolution window of . In step 1002, two spectral traces P are obtained by first obtaining an input spectral trace 1 (λ) and P 2 (λ):
[0140] P(λ)=h OSA (λ)*p(λ)=S(λ)+N(λ), (29)
[0141] Then with the convolution window h 1 (λ) and h 2 (λ) and integrate it to obtain P 1 (λ) and P 2 (λ) (see equations (25) and (26), where in this example h OSA (λ)=h OSA1 (λ)=h OSA2 (λ)). use h 1 (λ) Obtain two spectral traces P 1 (λ) and P 2 (λ), select h 1 (λ), so that in the immediate vicinity of the peak, that is, at λ p ±δλ (where δλ is greater than the acquisition resolution), S 1 The signal peak formed in (λ) is flat (eg Figure 11 shown), or S 1 The signal peak in (λ) exhibits a constant slope, ie ...
Embodiment 3
[0155] When the signal peak is almost as wide as the width of the noise component in the optical channel, Figure 10 Assumptions made in the method - at wider bandwidth RBW 2 The signal peak power integrated on is equal to when the RBW 1 Signal peak power integrated on - becomes unsuitable. However, when the spectral shape of the signal components is known, it is no longer necessary to select the first convolution window h that includes the full signal power 1 (λ). For example, the spectral shape of the signal component s(λ) can be derived by identifying the modulation spectrum of the signal, or dependent on the polarization properties as described above, which would be an embodiment of a hybrid approach as described below. The spectral shape of the signal component s(λ) can also be obtained by applying filtering techniques—assuming that the noise component n(λ) changes spectrally more slowly than the signal component s(λ) above the bandwidth of the optical signal. Compone...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More