CYCLOSTATIONARY FEATURES OF PAL TV AND WIRELESS MICROPHONE FOR COGNITIVE RADIO APPLICATIONS

Frequency spectrum being a scarce resource in communication system design,
spectrum sharing seems to be the solution to an optimal utilization of frequency
spectrum. The traditional fixed frequency allocation is not suitable for futuristic
networks that demand more and more spectrum for new wireless services. Cognitive
radio is a new emerging technology based on spectrum sharing concept.
Spectrum sensing is a vital task in this emerging technology by which it is able to
scan the frequency spectrum to identify the unused spectrum bands and utilize them.
In this thesis, we discuss spectrum sensing in the context of IEEE 802.22 Wireless
Regional Area Network (WRAN). In order to do so, we develop the co-existence
scenario with three cases according to geographical positions of primary services and
secondary service. In WRAN application, the SUs utilize the unused channel in TV
spectrum, which means that the primary users are TV service and other FCC part 74
low power licensed devices. We focus on special case of Analog TV-PAL service and
wireless microphone service as part 74 devices. Before discussing the spectrum
sensing technique, we propose architecture for sensing receiver. The concept of noise
uncertainty is also introduced in this context. The cyclostationarity theory is
introduced and we explain the motivation behind using the theory for spectrum
sensing and the reason that makes the cyclostationary features detector a powerful
detection technique in cognitive radio. We obtain the cyclostationary features of these
primary signals using spectral correlation function. Based on these features, we
develop two algorithms for spectrum sensing and their performances are evaluated in
comparison with energy detector which is considered as the standard simple detector.
Given that the cyclostationary features are unique for a particular signal; these
features can be used for signals classification. In our case, we use those features to
decide if the licensed channel is used by TV service or wireless microphone service.
This provides additional information for spectrum management and power control. Implementation issue is very important in cognitive radio generally and spectrum
sensing specially, hence we discuss the implementation of cyclostationary features
detector and compare its complexity with that of energy detector.