Blind Separation of Co-frequency Communications Recon Signals

Independent Component Analysis (ICA) has been widely used in the separation of mixtures of stochastic random signals. However, in the applications of radar systems and communication systems, we may encounter problems of ICA for deterministic signals, especially the sinusoidal signals. In this paper, we use kurtosis to indicate the independence of the signals and prove that the ICA can successfully separate the mixtures of deterministic sinusoidal signals. Which inspire us to apply ICA in complex signals (radar signals and communication signals). In the communication investigation field, the most popular application of ICA is to separate the mixtures of signals which are in the same frequency band. This paper proved that ICA can not only separate the co-frequency overlapping signals with different modulation, but also at a fast convergence rate and has good separating performance. This ICA technique can be used in the communication reconnaissance platforms based on the PXI.