Blind Signal Separation

Blind signal separation is known as a powerful tool for
enhancing noisy speech in many real world environments. In
this paper, the detail properties and structure of the newly
proposed null beamforming-blind separation (NBF-BSS)
algorithm is discussed. The performance of the blind signal
separation is improved using directional information of
sources. The beam patterns of the null beamformer, blind
signal separation, and the combination of the two are
investigated. It is shown that directional null reduces mainly
direct parts of the unwanted signals whereas blind source
separation reduces reverberant parts as well as direct parts.
Increasing the number of microphone sensors converge into
faster and improved separation by removing the direct parts
first by null beamforming. Simulation results using real office
recordings confirm the expectation.

Abstract, This paper proposes a new algorithm for solving the Blind Signal Separation (BSS) problem for convolutive mixing completely in the time domain. The closed form expressions used for first and second order optimization techniques derived in [1] are extended to accommodate the more practical convolutive mixing scenario. Traditionally convolutive BSS problems are solved in the frequency domain [2], [3],

Airborne antenna arrays suffer from the problem of multipath, formed due to scattering of the desired signal by the aircraft propellers. The direct path includes the desired signal coming from a sonobuoy floating in the ocean or from other airborne radars, and their reflections from the ocean surface. Numerous techniques have been proposed for multipath mitigation and localization of the

We present two novel methods for the blind extraction of multiple binary signals from a single observation generated from linear convolution and superposition. Both methods are based on two features: (a) the constellation structure of the successor set for each output value, and (b) the concept of system deflation, ie. the recursive shortening of the channel until it is reduced into an instantaneous mixture. The two methods differ in complexity as the latter requires significantly smaller amounts of data. Both methods are fast for small problem sizes but their complexities increase exponentially as the number of inputs and/or channel length increases. We present the complete analytical solution to the noiseless case and we thoroughly treat the related mathematical tools and concepts. The noisy case is briefly discussed as the complete treatment would exceed the scope of the present paper. 1.

The paper proposes a new technique for blind source separation (BSS) in the subband domain using an extended lapped transform (ELT) decomposition for nonstationary, convolutively mixed signals. As identified by S. Araki et al. (see Proc. 4th Int. Symp. on Independent Component Analysis and Blind Signal Separation - ICA2003, p.499-504, 2003), the motivation for subband-based BSS is the drawback of frequency domain BSS when dealing with separating mixed speech signals over a few seconds resulting in few samples in individual frequency bins leading to poor separation performance. In the proposed approach, mixed signals are decomposed into subband components by an ELT and within each subband a time domain Newton BSS algorithm is employed based on the nonstationarity property of the input signals and the joint diagonalization of output correlation matrices with time varying second order statistics (SOS). This subband version is compared to a fullband version using the same BSS algorithm.