Abstract

High data-rate WDM optical transmission network management requires monitoring a variety of channel performance parameters such as wavelength, power, SNR, etc. Cost-effective monitoring solutions aim to perform most of the processing electronically, in a bid to reduce the number of expensive optical components. The (spatial) independence between the transmitted WDM channels has been exploited in recent works.^1,2,3 The technique³ is able to reconstruct the complete channel waveforms, from which performance parameters can then be measured. Wavelength-dependent attenuators (WDAs) are employed to obtain additional observations of the WDM signal, each observation being considered as a mixture of the constituent channels. Because the independent channels contribute with different strengths to each observation, sufficient spatial diversity is available for a suitable blind signal separation (BSS) method to recover the original transmitted waveforms. The symmetric adaptive decorrelation (SAD)⁴ was adopted as a separation device. This particular technique, however, presents a number of deficiencies. Its complexity is of order O(N!) for an N-channel WDM transmission, and it has inherent stability and convergence difficulties—including spurious non-separating solutions⁴—which may hinder the monitoring process in practical cases. Also, the method is based on second-order statistics, which causes identifiability problems in the separation of spectrally white sources.

© 2002 Optical Society of America