The combined use of a Doppler radar with digital signal processing technique gives an effective non-invasive remote sensing of heart beat signals. Initial results have showed that the proposed technique is very promising in successfully isolating desired heart beat signals from other mobile objects and other distortion effects characterizing the wireless channel. We concentrate in this paper on the harder problem of separating at low SNR the signals from two subjects in the same room. We show that preliminary results obtained by the real analytical constant modulus algorithm (RACMA) and the independent component analysis (ICA) on experimental data are very promising into this goal.

We investigate the separation of a linear mixture of secondary surveillance radar (SSR) replies impinging on an M-element antenna array. At base-band, a received SSR signal consists of a binary sequence with alphabet {0, 1}, modulated by a complex exponential due to the residual carrier frequency. We present two algebraic algorithms to compute the separating beamformers by taking into account the particular modulation format of the received signal.

We prove that mixtures of continuous alphabet constant modulus sources can be identified with probability 1 with a finite number of samples (under noise-free conditions). This strengthens earlier results which only considered an infinite number of samples. The proof is based on the linearization technique of the analytical constant modulus algorithm (ACMA), together with a simple inductive argument. We then study the finite-alphabet case. In this case, we provide a subexponentially decaying upper bound on the probability of nonidentifiability for a finite number of samples. We show that under practical assumptions, this upper bound is tighter than the currently known bound. We then provide an improved exponentially decaying upper bound for the case of L-PSK signals (L is even).

The implementation of a RAKE receiver for a quasi-synchronous DS-CDMA (direct sequence code division multiple access) system requires the estimation of the dominant path delays for each user. We present a blind method to estimate these parameters. The algorithm takes advantage of various signal space invariances in the frequency domain to isolate the subspace of interest for each user, then uses ESPRIT on these subspaces to estimate the delays. The method processes each user independently of the others, is near-far resistant, and allows several delays per user. Simulations indicate a fair accuracy.

Unlike the Schetzen method, this identification method proposed can process inputs with discrete distribution, accepts kernels of any length, tolerates signal-independent zero-mean additive noise irrespective of its color and distribution, and leads to a closed form solution. Identifiability is proved under assumptions on input moments; more accurate conditions are derived for particular discrete inputs (in particular in the complex plane, distributions symmetric about both axes). Several simulations demonstrate its efficiency and its speed for various discrete input distributions.