A Fast Spectrum Sensing Method and Device Based on Binomial Distribution

Abstract

The present invention discloses a method and device for quickly sensing spectrum based on binomial distribution. The device comprises a sampler, a numerical calculator and a decision device, wherein the sampler is used for sampling a receiving signal, and transmitting a sampling point numerical value to the numerical calculator; the numerical calculator is used for calculating a sampling point signal and transmitting a calculating result to the decision device; and the decision device receives the calculating result of the numerical calculator for deciding. A calculating method for test statistics in the present invention is simple, does not involve multiplication operation, and only involves comparison operation and additive operation, thereby shortening the sensing time and facilitating real-time detection.

Description

【Technical field】 [0001] The present invention relates to the technical field of spectrum sensing in the field of cognitive radio, in particular to a fast spectrum sensing method and device based on binomial distribution. 【Background technique】 [0002] Cognitive radio (CR) technology is a dynamic spectrum management technology, which aims to solve the increasingly serious problems of scarcity of spectrum resources and low spectrum utilization. The core idea is to allow the secondary user (SU) to dynamically access the licensed frequency band when the primary user (PU) does not use the licensed frequency band, and when the primary user (PU) re-uses the licensed frequency band, the secondary user can withdraw in time to avoid interfering with the primary user's communication. It can be seen that the prerequisite and primary task of cognitive radio is spectrum sensing, that is, secondary users can quickly and accurately perceive whether there is a primary user signal in the fr...

AI Technical Summary

Problems solved by technology

[0003] Classical spectrum sensing methods mainly include cyclostationary feature detection methods, matched filter detection methods, energy detection methods, and sensing methods based on covariance matrix feature structures. Cyclostationary feature detection methods are highly complex, while matched filter detection methods must predict The user's prior knowledge (such as signal waveform, modulation method, etc.) and the requirements for synchronization are relatively high, and the practical application has great limitations. The energy detection (ED) method was introduced by Harry Urkowitz in the document "Energy detection of unknown deterministic" in 1967. Signals" ("Energy Detection Method for Deterministic Signals"), this method does not need to know any prior information of the main user, and the hardware implementation is simple, and only a small number of sampling points are needed to obtain better perceptual performance, but It needs to predict the noise variance and is sensitive to noise uncertainty. Noise uncertainty often causes a sharp decline in algorithm performance. The perception method based on the characteristic structure of the covariance matrix overcomes this problem. Correlation detection, mainly based on the ratio of the maximum and minimum eigenvalues ​​and the method based on feature template matching, etc., its detection performance is better than energy detection, does not need to predict any prior knowledge, and is not affected by noise uncertainty, But its disadvantage is that it needs more sampling points, and the computational complexity is high, so it is not suitable for real-time detection; recently, Shen Lei et al. published in the literature "Research on Spectrum Blind Detection Algorithm Based on Goodness of Fit in Cognitive Radio" in 2012 A spectral blind detection (CAD) method based on eigenfunction is proposed in the paper. This algorithm has good detection performance, does not need to know the noise variance and is not affected by noise uncertainty, but the disadvantage is that the detection method has a large amount of computation (test statistic The amount of calculation is large)

Images