Abstract

The wireless frequency spectrum is a very valuable resource in the field of communications. Over the years, different bands of the spectrum were licensed to various communications systems and
standards. As a result, most of the easily accessible parts of it ended up being theoretically occupied.
This made it somewhat difficult to accommodate new wireless technologies, especially with the rise of communications concepts such as the Machine to Machine (M2M) communications and the Internet of Things (IoT). It was necessary to find ways to make better use of wireless spectrum.
Cognitive Radio is one concept that came into the light to tackle the problem of spectrum utilization. Various technical reports stated that the spectrum is in fact under-utilized. Many frequency bands are
not heavily used over time, and some bands have low activity. Cognitive Radio (CR) Networks aim to exploit and opportunistically share the already licensed spectrum. The objective is to enable various kinds of communications while preserving the licensed parties' right to access the spectrum without interference.
Cognitive radio networks have more than one approach to spectrum sharing. In interweave spectrum sharing scheme, cognitive radio devices look for opportunities in the spectrum, in frequency and over time. Therefore, and to find these opportunities, they employ what is known as spectrum sensing. In a
spectrum sensing phase, the CR device scans certain parts of the spectrum to find the voids or white spaces in it. After that it exploits these voids to perform its data transmission, thus avoiding any
interference with the licensed users.
Spectrum sensing has various classifications and approaches. In this thesis, we will present a general review of the main spectrum sensing categories. Furthermore, we will discuss some of the techniques employed in each category including their respective advantages and disadvantages, in addition to
some of the research work associated with them.
Our focus will be on cooperative spectrum sensing, which is a popular research topic. In cooperative spectrum sensing, multiple CR devices collaborate in the spectrum sensing operation to enhance the performance in terms of detection accuracy. We will investigate the soft-information decision fusion approach in cooperative sensing. In this approach, the CR devices forward their spectrum sensing data to a central node, commonly known as a Fusion Center. At the fusion center, this data is combined to achieve a higher level of accuracy in determining the occupied parts and the empty parts of the spectrum while considering Rayleigh fading channels. Furthermore, we will address the issue of
high power consumption due to the sampling process of a wide-band of frequencies at the Nyquist rate. We will apply the 1-bit Quantization technique in our work to tackle this issue. The simulation results show that the detection accuracy of a 1-bit quantized system is equivalent to a non-quantized system with only 2 dB less in Signal-to-Noise Ratio (SNR). Finally, we will shed some light on multiple antenna spectrum sensing, and compare its performance to the cooperative sensing.