Optimizing the spectrum utilization in cognitive radio network through machine learning techniques

Abstract

The current IEEE 802.11 wireless networks suffer from many issues like speed, newlinesecurity, connectivity, network expansion, underutilization and wastage of Radio newlineFrequency (RF) spectrum. The increasing number of wireless users is the main cause newlineof underutilization or wastage of the RF spectrum. The researchers analyzed the newlineunderutilization of the RF spectrum and proposed a solution of using intelligent newlinewireless networks known as IEEE 802.22 Cognitive Radio Networks. The Cognitive newlineRadio Network (CRN) has two kinds of users: a) Primary user (PU) which has the newlineallocated license to use a particular frequency band of the RF spectrum b) Secondary newlineuser (SU) which do not have the license but uses the frequency band of the primary newlineusers when they are not using it. When the PU wants to use its allocated license band, newlineit preempts the SU. The secondary user of the CRN must have sensing, learning, newlineadapting and decision-making capabilities in order to detect the presence or absence of newlinethe PU in a particular frequency band of the RF spectrum. The SU sense the RF newlinespectrum to detect when the PU is not using its frequency band, so that it is available newlinefor their communication. The SU then learn the parameters of the frequency band of newlinethe RF spectrum and quickly adapt to the environment of the RF band. The secondary newlineuser then starts its communication over that particular frequency band of the RF newlinespectrum. When the PU preempts, the SU quickly sense the next available frequency newlineband of the RF spectrum, learn the parameters and adapt to the RF environment to start newlineits communication. This requires intelligence, decision making and switching newlinecapabilities. The secondary user of the CRN uses various spectrum sensing techniques newlinelike non cooperative spectrum sensing or cooperative spectrum sensing. In the noncooperative newlinespectrum sensing, the secondary users do not share their energy detection newlineinformation with others secondary users, whereas in cooperative spectrum sensing newline(CSS), the local secondary user receives the information of the energy fr