Faculty Advisor

Wyglinski, Alexander


This report describes the design and implementation of the spectrum sensing and signal classification sub-systems of a cooperative network. A sensor blindly receives and calculates the cyclic statistics of a signal decides whether or not the signal represents information or noise. If the signal's statistics indicate the presence of data, the system attempts to classify its modulation scheme. Finally, the decisions of several independent sensors are combined to provide a reliable estimate of the contents of the spectrum of interest. Independently, sensors correctly classify a signal about 60-70% of the time in a low SNR environment. The data fusion module improves this number significantly - especially as the number of sensors increases.


Worcester Polytechnic Institute

Date Accepted

December 2009


Electrical and Computer Engineering

Project Type

Major Qualifying Project



Advisor Department

Electrical and Computer Engineering

Your accessibility may vary due to other restrictions.