1-D and 2-D signals parallel processing in mixed formats

Paper presents general methods of processing one- and two-dimensional signals that are based on functional transform of signal using corresponding transformation functions, for example, as in DFT, Hilbert transform, wavelet transform, etc. Fast parallel algorithms for such functional transforms in complete (PCM-PCM) and mixed (ÑPCM-PCM or DM-PCM) formats are suggested. The latter ones are obtained using the proven theorems. In real-time mode and assuming the use of the same element base the suggested algorithms provide wider bandwidth of processed signals compared to those based on FFT. Their ability to provide results faster increases even more when substituting multiplication for shift operations. Examples of using the developed methods in parallel calculations structures are presented