User selection and density based decision method for cooperative spectrum sensing in cognitive radio networks

Abstract

Radio spectrum is one of the scarce source and its importance is increasing in recent years due to rapidly developing new technologies. Today's wireless networks use fixed spectrum assignment policy which causes inefficient utilization of this limited source. Therefore, new technologies are needed to use the spectrum in a more efficient manner. One of the promising technology that aims to use the spectrum dynamically is Cognitive Radios (CRs). Cognitive Radio Networks (CRNs) aim to use the spectrum dynamically by sensing the spectrum and if any gap is detected than they will provide their customer to use this spectrum band. Spectrum sensing plays significant role for the implementation of CRNs. The focus of this thesis is spectrum sensing problem in CRNs. In this work, spectrum sensing problem is investigated as a binary signal detection problem, which aims to find if a signal is present or not. The signal to be detected can be either known or unknown characteristics depending on the application. The signal is considered to be completely unknown from the Secondary User (SU) receiver side throughout the thesis. For this reason, energy detector is preferred as a detection method. Energy detectors (radiometers) are often used due to their implementation simplicity and good performances. The detection is based on some function of the received samples which is compared with a threshold. If the threshold is exceeded, it is decided that signal is present. Since sensing is very crucial at CRNs, cooperative user selection algorithms and decision methods are proposed to improve the performance of the detector. Two cooperative user selection algorithms based on SUs' and Primary User's (PU's) locations are proposed. Besides, two centralized cooperative decision methods are also proposed and evaluated with the cooperative user selection algorithms. One of the proposed cooperative decision methods is based on SUs' decision density at each region and the other one is based on weighting SUs' decisions according to their regions. Performance of these methods are evaluated using indoor and outdoor propagation models. One of the main goals here is to show that with the proposed cooperative selection algorithms, performance is improved when compared with ran- dom and worst selection of SUs. The other one is to illustrate further improvement in sensing performance is also achieved with the proposed centralized cooperative decision methods, compared to non-cooperative and cooperative schemes like majority rule and hard decision. The success of these methods are verified through simulations.