Wideband digital Quadrature Transformation based on perfect reconstruction filter banks

Quadrature Transformation based on Hilbert transform method to wideband signal is not suited to engineering application because the group delay can make severe aliasing. Wideband digital quadrature transformation based on perfect reconstruction filter banks is put forward in this paper. The wideband signal is divided into sub-band signals by analysis filter bank, then let the negative parts of the output sub-band signals equal to zero, then compose the processed signal by synthesis filter banks, the synthesis signal is the complex signal after quadrature transformation. Filter polyphase structure is used in this method, which can be realized by FFT algorithm. Algorithm of FFT and IFFT are used only one time in this new method which can reduce more computational workload, and it has the advantages of better image-reject, being realized easily in FPGA. The simulation result of Matlab shows that the error between the complex signals after quadrature transformation based on the traditional method and this paper´s method is up to -150dB, the error is small.