Hi and welcome to our WINLAB Summer project page!!

 

The Goal of our Project is to develop a generalized framework for spectrum sensing on USRP2 using GNU Radio.

 

 

Spectrum Sensing

 

Spectrum Sesnsing refers to the pratice of scanning the spectrum in search for vacancies in order to transform them into usable communication channels. Spectrum Sensing is an integral part of communications in the modern era due to the lack of spectral resources available today. There are numerous spectrums sensing techniques that have developed. However the 4 most commonly used techniques are as follows.

 

   1) Energy Detection

   2) Cyclostationary Method

   3) Matched Filter Detection

   4) Wavelet Detection

 

The technique used in this project was energy detection. Energy detection is the most common technique used because of its low computational and implementational complexity. In energy detection we scan and measure the enrgy of the available RF spectrum and compare it with a predefined threshold. If the energy measured is below the threshold value, we assume that the spectrum is available for use. However, if the energy measured is above the threshold values, we assume that the spectrum is already in use by another use.

However, energy detection also has its limitation. Energy detection is unreliable in low SNR areas because it is unable to distinguish between just noise and signal+noise. Also energy detection is unable to distinguish between a Primary User(PU) and a Secondary User(SU). The Energy detection Algorithm is illustrated in the flowchart below.