Spectral correlation is perhaps the most widely used characterization of the cyclostationarity property. The main reason is that the computational efficiency of the FFT can be harnessed to characterize the cyclostationarity of a given signal or data set in an efficient manner. And not just efficient, but with a reasonable total computational cost, so that one doesn’t have to wait *too* long for the result.

Just as the normal power spectrum is actually the power spectral density, or more accurately, the spectral density of time-averaged power (variance), the spectral correlation function is the spectral density of time-averaged correlation (covariance). What does this mean? Consider the following schematic showing two narrowband spectral components of an arbitrary signal:

The sequence of shaded rectangles on the left are meant to imply a time-series corresponding to the output of a bandpass filter centered at with bandwidth . Similarly, the sequence of shaded rectangles on the right imply a time-series corresponding to the output of a bandpass filter centered at with bandwidth .

Let’s call the first time-series and the second one . Since these time-series, or signals, are bandpass in general, if we attempt to measure their correlation we will get a small value if . However, if we downconvert each of them to baseband (zero frequency), we will obtain lowpass signals, and there is the possibility that these new signals are correlated to some degree.

As a limiting case, suppose each of the signals were noiseless sine waves. By the construction of the figure, their frequencies must be different if , and so their correlation will be zero. However, if each sine wave is perfectly downconverted to baseband, the resulting signals are simply complex-valued constants, and the correlation between the two constant time-series is high.

In general, the narrowband time-series and are not simple sine waves, but complicated random processes. But the correlation between separated spectral components–** spectral correlation**–is still a highly useful characterization of the signal for a large class of interesting signals.

The spectral components (the individual downconverted narrowband spectral components of the signal) are most often obtained through the use of the Fourier transform. As the transform of length slides along the signal, if produces a number of downconverted spectral components with approximate bandwidth . The two involved time-series of interest are then renamed as and . A measure of the spectral correlation is given by the limiting average of the *cyclic periodogram*, which is defined by

as the amount of processed data increases without bound, and then the spectral resolution () is allowed to decrease to zero,

The *limit spectral correlation function* we just wrote down is a time-smoothed (time-averaged) cyclic periodogram. But the limit function can also be obtained by frequency smoothing the cyclic periodogram

where is a unit-area pulse-like smoothing kernel (such as a rectangle). In (3), the symbol denotes convolution.

**The Significance of the Frequency A**

The spectral correlation function (SCF) is typically zero for almost all real numbers . Those for which the SCF is **not** identically zero are called ** cycle frequencies** (CFs). The set of SCF CFs is exactly the same as the set of cycle frequencies for the cyclic autocorrelation function (CAF)! That is, the separation between correlated narrowband signal components of is the same as a frequency of a sine wave that can be generated by a quadratic nonlinearity (for example, a squarer or a delay-and-multiply device) operating on the original time-series data .

**The Cyclic Wiener Relationship**

It can be shown that the Fourier transform of the CAF is equal to the SCF (The Literature [R1], My Papers [5,6]):

which is called the cyclic Wiener relationship. The Wiener relationship (sometimes called the Wiener-Khintchine theorem) is a name given to the familiar Fourier transform relation between the conventional power spectral density and the autocorrelation

where is the conventional power spectrum and is the conventional autocorrelation function.

**Conjugate Spectral Correlation**

This post has defined the **non-conjugate** spectral correlation function, which is the correlation between and . (The correlation between random variables and is defined as ; that is, **the standard correlation includes a conjugation**.)

The *conjugate SCF* is defined as the Fourier transform of the conjugate cyclic autocorrelation function,

From this definition, it can be shown that the conjugate SCF is the density of time-averaged correlation between and ,

where is the *conjugate cyclic periodogram*

The detailed explanation for why we need two kinds of spectral correlation functions (and, correspondingly, two kinds of cyclic autocorrelation functions) can be found in this post.

### Illustrations

The SCF below is estimated (more on that estimation in another post) from a simulated BPSK signal having bit rate of kHz and carrier frequency of kHz. A small amount of noise is added to the signal prior to SCF estimation. The (non-conjugate) SCF shows the power spectrum for and the bit-rate SCF for kHz. The conjugate SCF plot shows the prominent feature for the doubled-carrier cycle frequency kHz, and features offset from the doubled-carrier feature by kHz. More on the spectral correlation of the BPSK signal can be found here and here. The spectral correlation surfaces for a variety of communication signals can be found in this gallery post.

A closely related function called the spectral coherence function is useful for blindly detecting cycle frequencies exhibited by arbitrary data sets.

Now consider a similar signal: QPSK with rectangular pulses. Let’s switch to normalized frequencies here for convenience. The signal has a symbol rate of , a carrier frequency of , unit power, and a small amount of additive white Gaussian noise. A power spectrum estimate is shown in the following figure:

Consider also four distinct narrowband (NB) components of this QPSK signal as shown in the figure. The center frequencies are , , , and . We know that this signal has non-conjugate cycle frequencies that are equal to harmonics of the symbol rate, or for . This means that the NB components with separations are correlated. So if we extract such NB components “by hand” and calculate their correlation coefficients as a function of relative delay, we should see large results for the pairs and , and small results for all other pairs drawn from the four frequencies.

So let’s do that. We apply a simple Fourier-based ideal filter (ideal meaning rectangular pass band) with center frequency , frequency shift to complex baseband (zero center frequency) and decimate. The results are our narrowband signal components. Are they correlated when they should be and uncorrelated when they should be?

Here are the correlation-coefficient results:

Here the signals arise from the frequencies . So the spectral correlation concept is verified here. One can also simply plot the decimated shifted narrowband components and assess correlation visually:

I would like to know how I choose the parameters (time delay & cyclic frequency) to plot CAF in Matlab ?

Do you mean you want to plot the spectral correlation function? Or the cyclic autocorrelation?

I’m a bit confused because you mention the CAF but the comment is in response to the SCF post.

For either function, the cycle frequencies are, in the case of the post, the known cycle

frequencies for the BPSK signal with bit rate of 333.3 kHz and carrier offset frequency of 100 kHz.

This particular signal has a bandlimited pulse shape (not rectangular in the time domain,

as we usually use on this site), so the non-conjugate cycle frequencies are +333.3, 0, -333.3 kHz.

The conjugate cycle frequencies are 2×100 = 200 kHz, 200 – 333.3 kHz, and 200 + 333.3 kHz.

In your estimator, you need to use normalized frequencies at some point.

In my case, the sampling rate was 1 MHz, so that the normalized bit rate is 1/3, and

the normalized carrier offset is 0.1. That leads to normalized non-conjugate cycle

frequencies of -1/3, 0, 1/3 and normalized conjugate cycle frequencies of 0.2, 0.2-1/3,

and 0.2+1/3.

For the CAF, you can first create estimates of the SCF for each cycle frequency,

then inverse Fourier transform them. The SCF estimates will correspond to all spectral

frequencies between negative half the sampling rate and positive half the sampling rate,

and transforming will then provide you with a CAF estimate over many distinct values of

the lag variable. Using that method, you don’t have to choose any particular lags,

you get them all.

I would like to know the formula of the threshold or the average to make the decision about spectrum sensing ? (likely as energy detector algorithm for spectrum sensing )

I don’t have a formula for the threshold for detection based on spectral correlation. I typically choose a threshold using empirical methods, which can take into account all the real-world deviations from a typical simple AWGN model. However, if you use the spectral coherence function, there is a way to compute a threshold that is based on some older work on cross spectral density measurements. I have a post in the works on how to compute and apply that thresholding formula. For now, look at The Literature [G. Carter, R64].

Hey Chad, I just want to say that this post really helped me understand what the spectral correlation function is conceptually (the correlation between different FFT frequency bins as they change over time). That first diagram was especially helpful. Thanks.

You’re welcome! Thank

youfor taking the time to write. Comments like yours keep me going.Hi Chad. I think equation 8 lose a conjugate symbol. And can I regard X_T as complex envelope?

Eq (8) is correct as is. Notice that this is the

conjugatecyclic periodogram, not the normal one (which is Eq (1)). You might want to read the post on conjugation configurations, which also explains why we need to consider a variety of numbers of conjugated terms when we study higher-order statistics for complex-valued signals. Yes, is the complex envelope.I see. Thank you.

What dose U mean in equation 2?

It is a variable with units of time, like , and is equal to the time interval over which the time-varying cyclic periodogram is averaged.

Dear Chad, Thank you very much for your explanation. I followed several resources to understand the concept of spectral correlation, but those weren’t so helpful. Your post is the ideal.

Hey Chad,

Thanks for this valuable blog.

I have some difficulty to understand the physical meaning of the cyclic spectrum.

Let’s pick a specific point (f1,alpha1,a1) in the 3D spectral correlation figure, where f1 is the spectral frequency value, alpha1 is the cyclic frequency value, and a1 is the magnitude value. And Let’s say f1=10KHz and alpha1=100KHz. What these values are representing related to the original signal?

represents the complex-valued correlation between two narrowband frequency components. The first is at and the second is at (assuming we’re talking about the non-conjugate spectral correlation function for now). But it is really the density of correlation, meaning that the ideal spectral correlation function considers the two narrowband frequency components to have infinitesimal width.

So represents the “idealized” correlation between the time-varying fluctuations in a very narrow band centered at kHz and the time-varying fluctuations in a second very narrowband centered at kHz.

Does that help?

For a specific alpha, the correlation has to be between 2 narrowband frequency components or it can be between 2 shifted versions of the whole spectrum?

For a specific and , the correlation is between two narrowband components. But the spectral correlation function is a (typically continuous) function of and discrete function of , so if you compute the entire function, you get something that you might call a set of correlations between all possible shifts of the entire spectrum.

If you fix , and let vary over , then you’ll get a set of all the possible correlations between narrowband spectral components whose separation is .

If you fix , and let vary over , then you’ll get a set of all the possible correlations between narrowband spectral components whose center frequencies have average value .

OK that make sense, thanks very much