Definition
Digital Filter can be divided into two parts: a classic filter and a modern filter. The classic filter is assumed that the useful components in the input signal x (n) and the desired filter elimination component are respectively located in different frequency bands, so we can filter the noise by a linear system. If the noise, the spectrum of the signal is mixed, The classic filter does not receive the filtering requirements. Typically there is a high pass filter, a low pass filter, a bandpass filter, a blocking filter. The modern filter is an useful signal and noise signal from a signal containing noise. This method is to treat signal and noise itself as a random signal, using its statistical features, such as autocorrelation functions, interim function, self-power spectrometry, and utilize digital devices. There are mainly digital filters such as Wiwan filter, Kalman filter, adaptive filter.
Classic filter
The concept of classic filtering is based on a project concept proposed in Fourier analysis and transformation. According to the higher mathematical theory, any signal that satisfies a certain condition can be seen as an unlimited sinusoidal wave. In other words, the engineering signal is a sine wave linearly superimposed to different frequencies, and the sine wave of different frequencies of the constituent signal is called the frequency component of the signal or the harmonic component. In fact, all electronic systems have their own frequency band width (restrictions on the highest frequency of the signal), and the frequency characteristics reflect this basic feature of the electronic system. The filter is the engineering application circuit designated according to the effect of circuit parameters on the frequency band width.
Modern filter
Modern filter thinking is very different from classic filtered ideas. Modern filtering is the essence of the randomness of the signal, and the signal and its noise are used as a random signal, and the signal itself is estimated by utilizing statistical features. Once the signal is estimated, the resulting signal itself is much higher than the original signal-to-noise ratio. A typical digital filter has a means of Kalman filtering, Wenner filtering, adaptive filtering, wavelet transform (Wavelet). Essentially, digital filtering is actually an algorithm that is implemented on digital devices. The digital devices here include not only the computer, but also embedded devices such as DSP, FPGA, ARM, etc.
Digital filtering performance
Digital filter has high precision, high reliability, program-controlled change characteristics or multiplexing, easy integration. Digital filtering has been widely used in language signal processing, image signal processing, medical bob signaling, and other applications. Digital filter has low pass, high-pass, bandpass, and all-in-one types. It can be unchanged or time-change, causal or non-causal, linear or non-linear. The most widely used linear, when the constant digital filter
Other classification
from the airspace, can be divided into limited digital filtering FIR, unlimited digital filtering IIR and adaptive filtering .