Font Size: a A A

Study On Signal Parameters Estimation And Detection Based On Signal Phase Matching Principle

Posted on:2006-09-19Degree:MasterType:Thesis
Country:ChinaCandidate:L LiuFull Text:PDF
GTID:2132360152482113Subject:Underwater Acoustics
Abstract/Summary:PDF Full Text Request
Signal parameters estimation and detection is a main direction of underwater acoustic signal processing, its application is very widely. This dissertation is devoted to the study of noise suppression, sinusoid signal parameters estimation and detection.The Signal Phase Matching Principle of three sensors is proposed in this dissertation. By using this principle, the formulae of estimating signal are derived. The reasons of error produced are analyzed. The physical meaning and the geometric interpretation of these principles are given.The method of estimating sinusoid signal's parameters is presented by using the Signal Phase Matching Principle. The formulae of estimating sinusoid signal's parameters are derived on the condition of signal's frequency is known and frequency is unknown but the frequency band is known. This method can weaken the affection on the estimation performance effectively, which is caused by the difference of signal's phase and amplitude come from each sensor. The algorithm is improved to get answer by least squares method, this improvement can avoid 'singular peak', so make the estimation precision better. The factors that affect the algorithm's performance is analyzed, and it is given that theoretic optimal parameters and rule of parameters choosing in actually application.A method for detecting sinusoid signal by using Signal Phase Matching Principle is given. The formulae of estimating sinusoid signal's parameters are derived on the condition of signal's frequency is known and frequency is unknown but the frequency band is known. The simulation proves that its performance is as same as the General Likelihood Ratio Test, or better than it. The factors that affect the algorithm's performance are analyzed, and it is given that theoretic optimal parameters and rule of parameters choosing in actually application.A improved Signal Phase Matching Principle- Signal Phase Matching Principle based on autocorrelation is proposed in this dissertation. This principle is applied on parameters estimation and detection of sinusoid signal, theory and simulation are allprove that the performance is better. If the sinusoid signal's frequency is known or not, the perform of parameters estimation and detection based on this method is better than that based on Signal Phase Matching Principle and Maximum Likelihood Estimation (General Likelihood Ratio Test), even better than the matched filter on the condition of former. Detection of multiple signals is studied, the change of detection performance is analyzed what is contact with the number of sinusoid signals and the ratio of each amplitude.
Keywords/Search Tags:Signal Phase Matching Principle, parameters estimation of sinusoid signal, detection of sinusoid signal, least square, autocorrelation
PDF Full Text Request
Related items