Blind deconvolution of coloured signals based on higher-order cepstra and data fusion
A new nonparametric blind deconvolution algorithm for coloured non-Gaussian signals is presented. The goal of blind deconvolution is the reconstruction of the input of an unknown linear time-invariant (LTI) mixed-phase system based only on the system output. The existing deconvolution schemes that require the least amount of knowledge about the input signal and the LTI system were developed for white input signals, or rely on parametric modelling of both the system and the input. To develop nonparametric algorithms for the deconvolution of coloured processes of unknown statistics, we are forced to consider a two-channel approach. The proposed algorithm utilises the data collected by two different receivers, each being the output of a different system due to the same input. The two systems are then reconstructed combining higherorder statistics of the measured signals and the theory of signal reconstruction from the higherorder spectral phase only.