The Research On Spectrum Sensing Technologies For Cognitive Radios

Cognitive Radios have been proposed as a technology to implement opportunistic sharing since they are able to sense the spectral environment over a wide frequency band and exploit this information to opportunistically provide wireless links that meet the user communications requirements optimally.This paper discusses one of the key technologies in cognitive radio(spectrum sensing),mainly studies the technologies of square-law combining and cooperative spectrum sensing based on cooperative relay.1.We derive closed-form expressions of detection probability for square-law equal gain combining and selective combining over Nakagami channels.We then analyze the sensing performance,and achieve the following results:1) Both equal gain combining and selective combining can improve sensing performance effectively,and the performance improvement is significant if we employ more diversity branches.2) The sensing performance of equal gain combining is always better than that of selective combining.The performance difference is little when diversity order is small. However,there is significant sensing performance difference if diversity order is large.3) The larger the Nakagami parameter is,the better square-law combining works. Moreover,there is remarkable improvement of sensing performance when the Nakagami parameter varies from 1 to 2,while the improvement is much smaller with further increasing of the Nakagami parameter.4) Sensing performance of square-law combining over Nakagami channels is much better than that over Rayleigh channels if Nakagami parameter is larger than 2.2.We define user and network cooperative gain in a two-user network,and exhibit the relationship between user cooperative gain and network cooperative gain(and agility gain).We then give eight conclusions that reflect the impacts of power amplify factor of relay user on sensing performance,which can be used to explain the results of our numerical simulations.Moreover,we derive the expression of signal-to-noise ratio (SNR),and exploit the improvement on SNR to explain why the sensing performance may be improved by using cooperation.We also propose a novel numerical model for sensing performance analysis,in which factors neglected in traditional model are taken into consideration.Using the new model,we discuss impacts of different parameters on the improvement of sensing performance,and exhibit the condition under which the cooperative sensing technology based on cooperative relay is suitable,and show the important parameter(beta) that impacts the sensing performance significantly.Further,we find that the cooperative scheme based on cooperative relay is not suitable for the IEEE 802.22 systems.3.How to select a relay is very important in a multi-user network.Particularily, sensing performance of a cognitive user should be improved at least when it employs another user to act as its relay.We define a new notation gain point.The improvement of sensing performance can be attained if the received power of its relay from PU is larger than the gain point.Under the unconstrained scheme,we prove the range of gain point by using H(o|¨)lder inequality.Instead of solving intergral equation group,an approximated linear function for determining the gain point is presented.For the constrained scheme,by modifying the scaling factor of the unconstrained scheme,we find that sensing performance of the unconstrained scheme can act as an upper bound for that of constrained scheme as P₂＞P₁(P₁ and P₂ are the normalized power of primary signal at a cognitive user U₁ and its realy U₂,respectively).We then present the relationship of gain points between the constrained scheme and the unconstrained scheme.4.We take average power rather than instantaneous power as the decision statistic, and derive formulas of false alarm and detection probability in Rayleigh channels for cooperative spectrum sensing based on cooperative realy.We then analyze the improvement of sensing performance by using cooperation.Similar to the case where instantaneous power is taken as the decision statistic,the cooperative scheme is also suitable for the environment where primary user's transmission decays faster with distance than cognitive user's transmissions do.Also,if the transmission power of cognitive users is much smaller than that of the primary user or cognitive user's transmissions attenuate faster with distance than the primary transmissions do,it is difficult for a cognitive user to achieve the improvement of sensing performance by using cooperation.Moreover,beta is an important parameter that impacts the sensing performance significantly.Different from taking instantaneous power as the decision statistic,the improvement of sensing performance can be neglected if the received power of a cognitive user is larger than 3dB,and the cooperative scheme is mainly used to improve the detection ability of weak signals when the number of sample points is a little larger.