Low-voltage Low-power Switched-current Integrated Circuit Technology For Implementation Of Continuous Wavelet Transform

The Wavelet Transform is a new style mathematic analysis tool. It is a new theory system developed from the studies of Y.Meyer, S.Mallat and I.Daubechies in 1980s. The Wavelet Analysis is a kind of self-adapted time-frequency localized method, which tunes the time-frequency window automatically and has a strong flexibility. Wavelet transform has found a wide range of applications in the processing of signals due to its time-frequency localization characteristics. Because of the heavy computational cost, software implementations of wavelet transform cannot be used in real-time signal processing. Recently, some scientists have been researching in hardware implementations, but almost concentrating on the implementations of Discrete Wavelet Transform (DWT). Since the Continuous Wavelet Transform (CWT) is an effective tool for the analysis of non-stationary signals, and is better in data compression comparing with DWT, it has attracted much attention. At present, the method of realizing CWT is mainly in time-domain and frequency-domain. The former one has the merits of high speed and simple construct, while the latter one flexible to implement much more categories.The switched-current (SI) circuit reported by J.B.Hughes et al. in 1989 is the current domain sampled-data system. It has many advantages because current domain operation offers greater ease for high frequency operation, low voltage operation, and wide dynamic range. Simultaneously, the switched-current circuit is compatible with digital VLSI technology, which does not require the linear capacitors. And the switched-current integrator does not require the voltage operation amplifier. SI technology is one of the important implementation technologies of low-voltage low-power VLSI so far. Therefore, the method of implementing continuous wavelet transform using switched-current technology has a cheerful prospect.In this paper, the principles on VLSI realization of one-dimension CWT are expounded, and the relevant methods of the implementation are classified and compared with each other. And then, a novel method of implementing one-dimension continuous wavelet transform using switched-current circuit is proposed, in which the continuous wavelet transform is implemented by the parallel structure of biquadratic functions realized with switched-current circuits on the basis of approximation theory of network function. The Marr wavelet is selected as an example with the construction procedure and structure of switched-current circuits elaborated. The parallel structure...