Proactive spectrum sensing and spectrum sharing for cognitive radio

• Main Author: Ian Bajaj
• Other Authors: Gong Yi
• Format: Theses and Dissertations
• Language: English
• Published: 2012
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Wireless communication systems
• Online Access: https://hdl.handle.net/10356/47841
• Institution: Nanyang Technological University

Cognitive radio (CR) is emerging as a field essential to the progress of wireless communications. With spectral efficiency at the forefront in protocol design, CR aims in combating the supposed spectrum crisis due to inefficiency in spectral utilization. Thus a lot more spectrum-hungry services can be supported by the limited frequency bands through opportunistic access or sharing of licensed spectrum. As spectrum sensing precedes any possible access or sharing of spectrum, it is crucial that the sensing result is accurate when the initial sensing decision is made. Most of the current sensing techniques are passive in nature (i.e. they do not involve any participation of the CR) and hence can incur sensing errors when eavesdropping in the midst of high-noise or fading environments. Thus we consider a novel proactive spectrum sensing scheme that increases the certainty in the spectrum decision while combating the hidden-node problem. This is achieved when the CR sends out a probing signal prior to the sensing process. We study the results of applying power constraints to the power available for probing, and see how this scheme proves beneficial to the licensed user (LU) in terms of greater protection while still maintaining the throughput needs of the CR. After proactive sensing we focus our attention on the issues raised with spectrum sharing in the field of CR communication. As outlined by most work in literature, cross-channel estimation still remains an issue. Assumptions regarding cross-channel information are common amongst work that deals with a spectrum sharing protocol. In actuality this is impractical, and there is no way for the CR to ascertain the actual interference it deals to the LU. Thus we propose a supervised probing and sensing model that enables a CR to estimate the magnitude of the cross-channel link between itself and the LU, thus allowing for interference-shielding. With probing power playing an essential role in cross-channel estimation, we optimize the incremental probing step-size to allow for the highest estimation success and study its effect on cross-channel estimation error.