Blind separation and capture of communication signals using a multitarget constant modulus beamformer

A multitarget least-squares CMA (constant modulus algorithm) that blindly adapts a narrowband beamformer to separate and capture multiple communication signals by exploiting only the low modulus variation of the communication waveforms is presented. It is shown that the algorithm can detect and classify transient and nontransient communication signals on the basis of modulus variation and extract those signals with maximum attainable output SINR (signal-to-interference and -noise ratio) in environments containing strong transient or nontransient cochannel interference. This is demonstrated via computer simulation for a narrowband (25 kHz bandwidth) four-element antenna array in environments containing FSK (frequency shift keying) bursts received at low ( approximately 4 dB signal-to-white-noise ratio) power levels in the presence of severe ( approximately -50-dB signal-to-interference ratio) interference. It is shown that the algorithm can capture the FSK bursts in less than 2 ms, and maintain capture on the signals as other signals appear and disappear over the signal ON times.<<ETX>>