| We changed the way data is collected in analog beamforming phased arrays. By using non-traditional beam patterns and exploiting sparsity in images, we reduced imaging scan time as well as hardware complexity in phased array devices. This approach generates pseudorandom beam patterns and recovers the desired signal information using reconstruction algorithms. Sparsity is exploited by combining the techniques of phased array imaging with a new method called compressed sensing. By linearizing the phased array problem into that of a simple gain function, powerful existing basis pursuit techniques can be used to reconstruct images from highly subsampled data, using very limited hardware. Phase switching hardware can be reduced to just 2 or 3 phase paths per sensor, which can dramatically reduce the size of the beamforming network, as well as the bus width for the phase control. In our simulations we have been able to obtain perfect or near perfect reconstruction of the signal environment, under certain circumstances. In some cases these methods produce better results than the conventional phased array technique, as the reconstruction removes the effects of the beam pattern. This significantly reduces the cost and power of imaging devices without compromising the quality and resolution of the resulting images. |
