Analysis And Study On Properties Of Commonly Used Digital Signal Detection Based On Cyclic Spectrum Detection Method

In today’s rapid development of the information age,along with the rapid development of wireless tecnology,wireless business more and more abundant. Spectrum sensing,spectrum analysis and spectrum judgment are the three main steps of the cognitive radio system,spectrum sensing is the first important step,This detection technology includes matched—filter detection,energy detection and cyclostationaiy feature detection.Noise signal is stationary signal,so cyclostationary feature sensing technology compared with the other detection technologies can more accurately distinguish noise signal. Meanwhile,each modulation signal has its own cyclsstationary properties,so cyclostationary feature detection can identify modulation users more accurately than other detection methods.The content of this paper is composed of the following four aspects:Firstly, with regard to sensing spectrum techniques, the dissertation systematically investigated matched filtering, energy detection, cyclostationary feature detection.Secondly,the spectrum characteristics of the widely used digital signals(2ASKMPSKN16QAMn、FSK和MSK)has analysis. Based on this, the recognition of their modulation methods and process are introduced, and simulation was compared in recognition performance under different signal-to-noise ratio.Thirdly, according to the limited-storing ability of the monitoring node, the frequency smoothing method of the cyclic-spectrum estimation has been researched. And the affect of single f profile to parameter estimation has been analyzed, this article provided a kind search algorithm based on neighboring f profiles. In this way, the information of neighboring f profiles of spectrally smoothed cyclic periodogram can be fully used.Finally, the application of symbol rate estimation based on the neighboring f profiles search algorithm has been researched. According to the neighboring f profiles search algorithm, the symbol rate estimation algorithm of2ASK、BPSK、QPSK、16QAM and MSK signals have been provided. Compared to single profile search algorithm, the simulation showed that the former could commendably improve the parameter estimation accuracy.