Band signal power is called the power spectrum on. It may be displayed with the distribution of the frequency change in signal power in certain regions. The spectrum is similar to a signal curve, in the field of science, the power spectrum and spectrum has some connection, but still not the same, there is a difference between them.
Power spectral density
In physics, the wave form of the signal is generally represented, for example, electromagnetic waves, sound waves or random vibration. When the power spectrum density wave is multiplied by an appropriate coefficient obtained after carrying wave frequency per unit power, which is called the power spectral density (power spectral density, PSD) signals.
Power spectral density spectrum is a probabilistic method, is a measure of the mean square value of the random variable. Usually for random vibration analysis, continuous transient response can be described by a probability distribution function, i.e., the level of probability of a corresponding response occurs. Defining the power spectral density of the "power" within the band (mean square) power spectral density statistics structure random dynamic loads excitation response is a power spectral density - versus frequency values, wherein the power spectrum density may be displacement power spectral density, power spectral density velocity, acceleration, power spectral density, power spectral density and other forms of force. Mathematically, the power spectral density - the area under the curve is the relationship between frequency values mean square value, when the mean square value is equal to zero mean and the variance, i.e., in response to a square value of the standard deviation.
While not necessarily a signal or its variable impart certain physical dimension, the following discussion is assumed that a signal change in the time domain. Defined above the energy spectral density Fourier transformation request signal must be present, that is square or square integrable signal can be added. Often a more useful alternative representation is the power spectral density (PSD), which defines how the time-series signal or power distribution over frequency. Here the power may be the actual physical power, or more often facilitate represent abstract signal is defined as the square of the signal value, i.e. when the load signal is the actual power at 1 ohm (ohm).
Since the average value is not zero signal is not square integrable, so in this case there is no Fourier transform. Fortunately Wiener - Khintchine Theorem (Wiener-Khinchin theorem) provides a simple alternative method, if the signal can be seen as stationary stochastic process, it is the signal power spectral density function from the Fourier-related transform.
the difference
Calculating a power spectrum needs to do first autocorrelation signal before FFT operation.
Computing sucked directly FFT spectrum signal on the line.
The signal power spectrum is studied, but it is in terms of energy research to the signal.
The spectrum is also used to describe the signal, but the representation has changed, from the time domain transformed into the frequency domain representation, that is to say it represents a different way of signals.
Defining the power spectrum in the case of the limited signal, converting conditions within the band range of the signal power, the power varies with frequency, so that performance as a power spectrum, which is the energy devoted to the power of the energy available is limited signals are analyzed exhibited . It contains some information on the magnitude of the spectrum, but the phase information is discarded out.
In contrast, the spectrum is not very strict, mainly reflects the average conversion signal, requiring only a period of time the average amount.
We often say that under different circumstances spectrum signal, its power spectrum is likely to be the same.
Although the power spectrum process is random, but because of the concept of statistics is average, the equivalent of stationary random process, power spectrum of this process is a deterministic function.
And the Fourier Transform of the sample spectrum, although the process is random, but random variations to this process, the formation of the spectrum of frequency domain sequence is random, function uncertain.
The power spectrum of the power spectrum and the concept of extremely magnitude difference is there, and their presence requirements are different. The existence of the power spectrum of the convergence requirements change, and only requires the existence of a spectrum of whether convergence.
Power spectrum and spectrum have the same place, and have ties, but these differences is important to determine their place two useful. Although power spectrum and spectrum signals are studies, but studies in different directions, the angle is not the same, and there are different from their nature, the power spectrum of the random bit worse, more stringent, there is determined a function of a support; and requires less spectrum, some randomness is quite strong, resulting in a change of its signal,