Replace PFB with WOLA

README.md

@@ -14,20 +14,12 @@ A list of GRC-supported SDRs can be found [here](https://wiki.gnuradio.org/index
 ![alt text](https://i.imgur.com/LOeADlg.png "Example Measurement")
 
 ## GRC Data Acquisition Flowgraph
-**CygnusRFI** is a [**polyphase filterbank** spectrometer](https://arxiv.org/abs/1607.03579). The raw I/Q samples are processed in real time using GNU Radio, with the amount of data stored to file being drastically reduced for further analysis. The following flowgraph handles the acquisition and early-stage processing of the data:
+**CygnusRFI** is a 4-tap WOLA spectrometer. The raw I/Q samples are processed in real time using GNU Radio, with the amount of data stored to file being drastically reduced for further analysis. The following flowgraph handles the acquisition and early-stage processing of the data:
 
 ![alt text](https://i.imgur.com/2Xp8qnZ.png "Data Acquisition Flowgraph")
 
-## Spectral leakage: a comparison between ACS, FTF and PFB spectrometers
-The noteworthy advantage of polyphase filterbanks is **reduced spectral leakage**, with a slight increase in computational requirements. The following figure compares the spectral leakage produced by an autocorrelation spectrometer (ACS), a Fourier transform filterbank spectrometer (FTF) and a polyphase filterbank spectrometer (PFB) with a Hann FFT window:
-![alt text](https://i.imgur.com/e5TwE3w.png "Spectrometer comparison regarding spectral leakage")
-Source: [Danny C. Price (2018)](https://arxiv.org/abs/1607.03579)
+Although not strictly necessary, in this application, minimal spectral leakage allows us to more effectively suppress sidelobes that are unintentionally produced next to the actual signal by the simple FFT algorithm. If for any reason WOLA is not used by the user (e.g. due to potential computational expense), it is suggested to at least use a window function when computing the FFT spectra, in order to suppress the generated FFT sidelobes to a somewhat acceptable level.
 
-Although not strictly necessary, in this application, minimal spectral leakage allows us to more effectively suppress sidelobes that are unintentionally produced next to the actual signal by the traditional FFT algorithm. If for any reason PFB is not used by the user (e.g. due to potential computational expense), it is suggested to at least use a window function when computing the FFT spectra, in order to suppress the generated sidelobes to a somewhat acceptable level.
-
-## A graphical representation of a polyphase filterbank
-![alt text](https://i.imgur.com/HUFTmTh.png)
-Source: [Danny C. Price (2018)](https://arxiv.org/abs/1607.03579)
 
 ## Data Analysis
 Once a single spectrum observation is finished and the data has been acquired and stored to `X.dat`, the FFT samples (interpreted as a [numpy array](https://docs.scipy.org/doc/numpy/reference/generated/numpy.array.html) in `rfi_plotter.py`) constitute the **dynamic spectrum (waterfall)**, from which the **averaged spectrum** of the observation can be derived.