Полосовая выборка

A bandpass signal has most of its energy concentrated within a narrow frequency band, with a spectrum that is nonzero only between defined lower and upper frequency limits. This characteristic makes bandpass signals important because they allow for more efficient sampling techniques compared to general signals, reducing the required sampling rate and data storage while maintaining signal integrity.

Multiplying the time-domain signal by an impulse train produces a sampled signal whose spectrum is obtained by convolving the original signal's spectrum with the impulse train spectrum over 2π. This convolution causes the signal's spectrum to repeat periodically with a period of 2π/T, where T is the sampling interval, creating multiple copies of the original spectrum at regular frequency intervals.

As the sampling interval T increases, the spacing between the repeated spectral copies decreases. This reduced spacing increases the risk of spectral overlap, which can lead to aliasing if the spacing becomes smaller than the signal's bandwidth, making careful selection of T critical for preventing signal distortion.

Aliasing is prevented by ensuring the spacing between repeated spectra is greater than the signal's bandwidth and by selecting the maximum allowable value of T accordingly. Additionally, filters with specific constants and frequencies are employed to isolate and pass only the frequencies within the desired narrow band, preserving signal integrity.

Bandpass sampling requires a sampling rate greater than twice the signal's bandwidth, not twice the highest frequency as in general sampling. This principle allows for lower sampling rates when the signal occupies a narrow frequency band away from zero, making bandpass sampling more efficient than Nyquist sampling for signals with concentrated spectral energy.

Convolution is used because multiplying a signal by an impulse train in the time domain corresponds to convolving their spectra in the frequency domain. This mathematical operation reveals how the original signal's spectrum repeats periodically after sampling, allowing engineers to analyze and predict aliasing conditions and design appropriate filters.

Filters with specific constants and frequencies pass only the frequencies within the desired narrow band, isolating the signal of interest from unwanted spectral copies. This selective filtering preserves the original signal's integrity during sampling and enables accurate reconstruction signal using interpolation techniques without distortion or aliasing artifacts.