Eurecom - Mobile Communications
Phd Student ( 2010 - 2013)
ThesisContributions to Spectrum Sensing and Cognitive Radio Systems
Cognitive radio (CR) is a smart wireless communication concept that is able to promote the efficiency of the spectrum usage by exploiting the free frequency bands in the spectrum, namely spectrum holes. Spectrum sensing is a major ingredient enabling CR technology, as it provides a series of algorithms enabling primary users (PUs) detection and spectrum holes identification.
In a first part, we study two spectrum sensing algorithms in wideband scenarios. The first algorithm is a reconfigurable filter bank that tracks discontinuities in PU transmissions due to transitions from occupied to free bands. Thus this detector enables a better identification of secondary user (SU) opportunities. The second contribution tackles a hardware related problem in wideband systems which is the unability of ADCs to acquire wideband signals at the Nyquist rate. We take into account this data acquisition problem by designing a spectrum sensing algorithm based on compressed sensing.
In the second part we focus on the identification of signals and standards in the TV White Space context. We first suggest a mixed signal separation and classification method based on multiple sensing algorithms. The second algorithm is directly detects cyclostationary features (CFD) of digitally modulated signals in order to classify PU signals (DVB-T, PMSE) while a SU (LTE) is transmitting. Here we also suggest to improve CFD performance via Computer Vision tools.
The third part proposes a network discovery mechanism (PU transmitter localization and spectrum sensing) in a unique compressed sensing framework. This framework exploits sparsity in two ways: sparsity of primary spectrum occupancy (which allows to overcome ADC sampling frequency limitations) and the fact that in a wireless networks transmitters are sparsely distributed within a given area.