Stable Distribution Noise Cyclostationary Signal Carrier Frequency Estimation

Signal detection and parameter estimation has always been the two most basic questions in the field of signal processing. In communications, telemetry, radar and sonar applications,many commonly used artificial signal is a special class of non-stationary signals, their specific order statistical parameters change with the time period, such signals are called "loop stationary signals. "In addition to the artificial signal outside the signal has some natural circadian or seasonal changes in the law, such as watermarks data, meteorological data, and other marine signal is cyclostationary signals. Cyclostationary signal processing is a new frontier of modern signal processing, there are many different from the traditional non-stationary signal processing theory and method. Using the signal characteristics of cyclostationary signal processing of the many traditional methods do not have the advantage.On the practical applications and theoretical study found such as radar, wireless communication systems often contain such a noise, it and our previous studies Gaussian noise are very different, which contains a large number of pulse ingredients, this noise is more suitable for use a stable distribution noise model to represent, so many of the traditional parameters of the proposed model based on Gaussian noise cyclostationary signal estimation theory can not effectively suppress the influence of impulse noise, thus affecting the validity and accuracy of parameter estimation and must therefore be studied to obtain a sufficiently suppress Gaussian noise, impulse noise signal processing new methods. So to achieve in complex environments containing stable distribution noise low noise carrier frequency signal noise ratio is estimated to be focused on the direction of signal processing. Matched filter requires knowing a priori information signal, but at low SNR or strong interference, a sharp decline in energy detection performance can not effectively estimate the parameters. However, cyclostationary feature detection method based on the above method can overcome the disadvantages of having a strong anti-noise performance, and can signal parameters under unknown circumstances prior information valid estimate signal carrier frequency, which work for the subsequent signal processing, such as symbol rate estimation, signal recovery and lay a solid foundation.