Low-complexity storage-reduced digital spectrum-based soft-failure management with Welch’s method

Abstract

In this paper, we propose a low-complexity storage-reduced digital spectrum-based soft failure management solution including soft failure detection (SFD), identification (SFI) and failure magnitude estimation (FME). Five soft failures are considered. Random fluctuation of key link parameters is introduced in simulations to investigate the robustness of the proposed solution. To reduce computational complexity and storage requirement for digital spectrum calculation, Welch's method is employed instead of applying fast Fourier transform on a long sampling sequence. A false positive rate below 1% and a false negative rate below 3% are achieved for SFD after extensive numerical simulations. Besides, SFI realizes an identification accuracy of 97.4%. The mean square errors of FME for different soft failures are all below 0.4. Finally, with a very small segment length of 25, 46.2% of computational complexity and 99.6% of storage can be reduced for digital spectrum calculation.