A fast sensing algorithm for spectrum detection in cognitive radios

• Main Author: Prashob Ramachandran Nair
• Other Authors: Vinod Achutavarrier Prasad
• Format: Theses and Dissertations
• Language: English
• Published: 2012
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Wireless communication systems
• Online Access: https://hdl.handle.net/10356/48179
• Institution: Nanyang Technological University

Cognitive Radio (CR) has emerged as a very promising solution to mitigate the scarcity in electromagnetic spectrum for wireless communication. CR promises to maximize the efficiency of spectrum utilization by seeking out spectral vacancies (holes) in the licensed band and using it for communication purposes. Thus sensing of the spectrum to detect the vacant holes forms an integral part of a CR system. The performance of a cognitive radio system depends on how well it can maximize the usage of a vacant frequency band without causing any interference to the licensed user, also called the primary user. Thus the ability to detect a primary user as well as to avoid any false alarm is of paramount importance for such a system. There are various techniques to examine the presence of a communication signal in a given frequency band. The easiest and most widespread of these is the energy based detection. In this technique, the energy of the received samples is computed and compared with a predetermined threshold. If the total energy of the received samples is more than the computed threshold, the frequency band is assumed to be occupied by the licensed user. The selection of the threshold to discriminate between the binary hypothesis of presence and absence of primary user signal plays a very vital role in determining the accuracy of the spectrum sensing algorithms. Conventionally, for an energy detector (ED) based CR, the threshold is set based on estimated noise power and this is not changed even though the CR may have more information available about the channel. It has been observed that for a conventional energy detector based CR at low signal-to-noise ratio (SNR), when the time to sense the spectrum is short, the ability to detect the presence of the primary user deteriorates significantly. However, if the threshold is chosen based on the target probability of detection, then there is a significant improvement in the ability of the CR to detect the presence of the primary user.