Selective discrete wavelet packet transform based energy detector for cognitive radios

Abstract

As growing the requirement of wireless communication service currently, the demand of frequency resource is increasing. Cognitive Radio(CR) technique was proposed to solve this frequency scarcity problem. Indeed, it makes use of wasted spectrum not to damage existing spectrum right. To utilize the unused spectrum safely, CR performs spectrum sensing, spectrum decision, spectrum sharing and spectrum mobility. Spectrum sensing makes a decision vacant frequency segment that is considered as the absence of licensed users. Spectrum decision determines the best tractable frequency region, and spectrum sharing adjusts approach to this part by sharing information with other users. When licensed users are sensed, spectrum mobility is performed to empty the frequency channels. In this thesis, we study the requirement of spectrum sensing and various spectrum sensing techniques. And also we propose the efficient spectrum sensing method to realize CR. To increase spectrum efficiency, IEEE 802.22 Working Group was established to apply CR to real TV spectrum. In this standard organization, efficient two stage sensing architecture was proposed to achieve accurate and fast spectrum sensing. It combines advantages of energy detector for fast detection and feature detector for fine detection. In this thesis, we purpose to realize more efficiency energy detector in two stage sensing architecture. As the energy detector that is first stage of two stage sensing architecture, DWPT (Discrete Wavelet Packet Transform) based energy detector was proposed. In this structure, energy detector carries out coarse sensing to find high power signals in wide spectrum band and feature detector performs fine sensing for low power signals in remaining spectrum band by using single threshold at first stage. And also Butter worth filters are implemented to realize DWPT for dividing frequency bands. In this method, however, load on feature detector increases due to many transmitted channels in ene...