For application requiring low voltage, low power and real time operating, approximation implementations of complex wavelet transform (CWT) in analog hardware can be achieved with sampled data circuits. A novel analog ...
详细信息
For application requiring low voltage, low power and real time operating, approximation implementations of complex wavelet transform (CWT) in analog hardware can be achieved with sampled data circuits. A novel analog implementation of CWT using adaptive chaotic differential evolution algorithm (ACDE) and switched-current (SI) filters is presented in this paper. The quality of such implementations depends on the accuracy of the corresponding wavelet approximations, so an optimized procedure based on ACDE is applied to approximate the real and the imaginary parts of the required complex wavelet. The global optimum parameters of the designed approximation model for the respective parts of complex wavelet function are solved. Then the respective approximationfunctions and transfer functions of complex wavelet filters are obtained. Finally, the CWT is implemented by the analog filter banks with SI integrators as main building blocks whose impulse response is the approximationfunction and its extensions of the real and imaginary parts. Simulations demonstrate that the proposed method is effectively to the CWT implementation.
暂无评论