This paper proposes a temporal sampling strategy that increases the accuracy of long-term noise level estimation and allows to establish the estimation error according to the number of sampled days. Days of the week are stratified into working days and weekend days. This research shows how to use measurements of L on working days to estimate the corresponding values for weekend days. This is possible because working days have higher noise levels and less variability than weekend days. The improvement in accuracy allows for a reduction in the number of required sampled days compared to taking samples randomly, which would help to reduce the uncertainty in environmental noise assessment. As a reference, to obtain a 90% confidence interval of ±1 dB for L, the proposed sampling strategy reduces the required measurement days by more than 38%. For L, the reduction is close to 18% of the total number of days. The proposed strategy could be adapted to different environments by simply changing a few parameters.

No abstract

No abstract

We propose and numerically analyze a colorless transmitter based on a Fabry-Perot Laser Diode in a WDM-PON. We study the power requirements for achieving a SMSR around 30 dB for different injected carrier linewidths.

No abstract