Performance comparison of time-domain and frequency-domain beamforming techniques for sensor array processing

This paper aims to provide performance comparison of various time-domain and frequency-domain beamforming techniques in terms of their execution time on general purpose processors along with the hardware considerations and their processing requirements. MATLAB simulations have been developed for timedomain beamforming techniques that include delay-and-sum, interpolation and quadrature beamforming. Similarly MATLAB simulations have been developed for frequency-domain beamforming techniques that include single-beam and multi-beam beamforming. High resolution beamforming techniques are used to obtain better spatial resolution for sensor arrays in comparison to their beam width. MATLAB simulations have been developed for high resolution beamforming techniques that include MUSIC and Root-MUSIC beamforming. The parameters used in these simulations include number of sensors (M), sensor spacing (d), array frequency (f), number of time samples (N), and sampling rate (fs). This paper shows that frequency-domain beamforming techniques are suitable for sensor arrays in terms of their processing requirements. The simulation results presented in this paper show that frequency-domain beamforming involving FFT and IFFT algorithms can be implemented with ease on general purpose processors and take less execution time as compared to other beamforming techniques which is an essential requirement for real-time systems.