18.1
משפט הדגימה
בעיבוד אותות, ניתוח של אותות רציפים בזמן, המסומנים כ-(x(t, כרוך לעיתים קרובות בטכניקות דגימה על מנת להמיר את האותות הללו לאותות בדידים בזמן. תהליך זה…
Consider the continuous-time signal x(t), and a train of impulses where Ts, is the sampling interval, and fs, is the sampling frequency.
Multiplying both signals results in a series of discrete impulses.
The Fourier transform shows that the spectrum of the sampled signal is a sum of shifted versions of the original signal's spectrum.
The spacing of these shifted versions is determined by the sampling frequency.
If the sampling frequency is greater than twice the highest frequency present in the original signal, these shifted spectra will not overlap.
This non-overlapping condition is crucial for the perfect reconstruction of the original signal from its samples.
The Sampling Theorem states that for a band-limited signal, the sampling frequency must be at least twice the highest frequency in the signal.
This minimum required frequency is known as the Nyquist rate, and meeting this criterion ensures no loss of information during the sampling process.
A signal is considered oversampled if sampled at a rate greater than its Nyquist rate and under-sampled if at a rate lower than the Nyquist rate.
View the full transcript and gain access to JoVE Core videos
Q1: What is the Nyquist rate and why is it important in signal sampling?
The Nyquist rate is the minimum sampling frequency required to perfectly reconstruct a band-limited signal. It must be at least twice the highest frequency present in the original signal. Meeting this criterion ensures no information loss during sampling and prevents spectral overlap in the frequency domain.
Q2: How does multiplying a continuous time signal by an impulse train create a sampled signal?
Multiplying a continuous time signal by an impulse train produces a series of discrete impulses at specific time intervals determined by the sampling interval. This operation converts the continuous signal into discrete samples that can be analyzed and processed digitally using the Fourier transform.
Q3: What happens to the frequency spectrum when a signal is sampled?
The Fourier transform reveals that the sampled signal's spectrum consists of multiple shifted versions of the original signal's spectrum, spaced apart by the sampling frequency. If the sampling frequency exceeds twice the highest frequency, these shifted spectra do not overlap, enabling perfect signal reconstruction.
Q4: What is the difference between oversampling and undersampling?
Oversampling occurs when the sampling frequency exceeds the Nyquist rate, providing benefits like reduced noise and simpler filter design. Undersampling happens when the sampling rate falls below the Nyquist rate, causing aliasing where different frequency components become indistinguishable, distorting the reconstructed signal.
Q5: How does the Sampling Theorem ensure perfect signal reconstruction?
The Sampling Theorem guarantees that when sampling frequency meets or exceeds twice the highest signal frequency, the spectral copies do not overlap. This non-overlapping condition is crucial for perfect reconstruction of signal using interpolation from discrete samples without information loss.
Q6: Why must the sampling frequency be greater than twice the highest signal frequency?
If sampling frequency is less than twice the highest frequency, shifted spectral versions overlap, causing information loss and aliasing. This overlap makes it impossible to distinguish original frequency components, corrupting the signal. The 2x requirement ensures spectral separation and accurate digital representation.
Q7: What role does the impulse train play in the sampling process?
The impulse train, characterized by the sampling interval and sampling frequency, acts as a mathematical tool to extract discrete samples from a continuous signal. When multiplied with the original signal, it creates discrete impulses at regular intervals, forming the foundation for digital signal processing and frequency domain analysis.