STRENGTHS AND WEAKNESSES OF CALCULATING BEAMFORMING IN THE TIME DOMAIN
Olaf Jaeckel · 2006
Most of the present beamforming systems in commercial and academic applications are performing their calculations solely in the frequency domain. While the early historical beamformers actually have been implemented in the time domain by inclusion of simple analog hardware delay lines, this principle was not applied anymore due to the availability of digital signal processors which allowed the calculation of fast Fourier transforms already within the data acquisition hardware frontends. Nowadays, modern computers offer the feasibility to process huge amounts of multichannel data in the frequency domain as well as in the time domain, so computational effort is not the only concern anymore. The paper compares the benefits and the disadvantages of both domains, focusing on the time domain because the latter is neglected or at least heavily underrepresented throughout the classical array signal processing textbooks and most of the current papers dealing with the beamforming method. Therefore, it is not the purpose of this paper to present a new theory or any advanced algorithms, but rather to give a general overview which is motivated from an application-oriented and practical standpoint of view. It is pointed out that a high signal bandwidth and a sampling rate much higher than twice the Nyquist frequency necessary for the channel data are especially important when working in the time domain. Application examples from technical and nontechnical fields will be given.