Research On Spectrum Sensing Algorithm Of Cognitive Radio Based On Welch Spectrual Estimation

Cognitive radio technology provide a viable idea for easing the contradiction between the lack of spectrum resources and growing demand for wireless access, by the "secondary use" approach of the authorized spectrum. In this way, spectrum sensing technology is considered the key to the realization of cognitive radio.IEEE802.22is the first standardization based on the cognitive radio technology in the world, which aims to realize Wireless Regional Area Network (WRAN) from point to multipoint, by treating the VHF/UHF bands assigned to the television broadcast as broadband chain by cognitive radio technology.This paper researches in spectrum sensing technology of cognitive radio, based on the perceptual framework of WRAN systems. WRAN spectrum sensing scheme is generally divided into two steps:first is coarse perception based on energy detection, we can identify that cognitive radio users could access the candidate channel initially; Next is fine features perception to the signal in the candidate channel, we can determine whether there is an authorized user in the channel.For channel coarse perception,, narrowband authorized users is easy to be drown in the noise in broadband VHF/UHF channels when SNR is low, which results in a sharp decline in the performance of energy detection algorithm. To address this problem, this paper put forward a frequency-domain energy detection algorithm based on Welch innovatively. This algorithm can obtain signal energy to achieve energy detection in frequency domain from the power spectrum in Welch spectral estimation. The simulation shows that the detect performance of this algorithm is better than traditional energy detection in time-domain with low SNR. Meanwhile, this paper analyzes the impact of noise fluctuations on this algorithm performance, and establishes a mathematical model of optimization decision threshold by the theoretical analysis to weaken noise. The Simulation results show that the proposed optimization threshold is better than the original threshold to improve the detection performance of the algorithm.For the fine features perception of the signal, this paper, establishes the cyclic spectrum model for different types of modulated signals by cyclostationary and smooth detection algorithm. The simulation verifies the differences of cyclostationary and smooth characteristic in different types of modulated signal, which shows that you can do fine features perception to signal by a smooth loop detection algorithm in WRAN system.