Energy based spectrum sensing for enabling dynamic spectrum access in cognitive radios.

摘要

Spectrum scarcity is increasingly becoming an obstacle for the implementation of new wireless technologies. On the contrary, recent studies have discovered considerable under-utilization of the allocated spectrum by the licensed users. This suggests that the solution to the problem is a transition from static spectrum allocation policies to dynamic spectrum access methodologies. This can be accomplished through the use of Cognitive Radio technology. Cognitive Radio is considered as an intelligent radio which is capable of altering its transmission or reception parameters in accordance to the radio environment and the network state to use the available spectrum optimally. Significant research efforts have furthered Cognitive Radios since the idea was first conceived by Joseph Mitola in 1998.;Cognitive Radio technology allows for the licensed spectrum of the primary users to be used on an opportunistic basis by unlicensed secondary users. A vital requirement of such an opportunistic scheme is that the licensed primary users be protected from detrimental interference from the secondary users while at the same time optimizing the performance for the secondary users. Thus the reliable detection of primary users offers better secondary system throughput via increased spectral efficiency in addition to safeguarding the primary system.;Spectrum sensing is a technique used to detect the presence of primary users in the licensed spectrum. It is the estimation of the instantaneous occupancy of the frequency spectrum and is a key enabling factor for Cognitive Radios. Various techniques exist for performing spectrum sensing. In addition to primary user detection, spectrum sensing can also be employed for secondary detection and co-existence, interference analysis in multi-radio environments etc.;In this thesis we study adaptive spectrum sensing based on energy detection with a purpose of demonstrating Dynamic Spectrum Access. The major focus has been to evaluate algorithms that can allow for estimation of noise in the presence of the signal which is essential for energy detection based schemes. We also present the system level implementation and evaluation of a Dynamic Spectrum Access setup developed using the USRP2/GNU Radio platform on the ORBIT Wireless Testbed at WINLAB as part of the WINLAB-NEC Collaborative Cognitive Radio Project.