Study On Signal Recognition And Parameter Extraction Of Pseudo-Code Compound System Fuze

In this dissertation, the recognition and parameter estimation of pseudo noise (PN)code compound systemic fuze are studied. The productions obtained from this dissertationare worth reference to the electronic reconnaissance and counter measures of PN systems,such as communication and radar. The main contributions are as followings:(1) The theory of adaptive window length time-frequency analysis and its applicationin the extraction of the time-frequency characteristic on pseudonoise code fuze signal areresearched. First, the general theory of adaptive window length time-frequency analysis isreviewed. Then two different algorithms, namely the WD distribution and the improved Bdistribution based on adaptive window length, are introduced and used respectively toextract the time-frequency characteristic of consecutive pseudonoise code fuze signals.Simulation results show that this method can not only extract the high frequencyinformation induced by the PN phase modulation well but also decrease the noise effect onextraction of instantaneous frequency. The parameters of the modulated signal areextracted according to the time-frequency characteristics.(2) An extraction method of pseudonoise code is given according to the instantaneousmaximum output amplitude of fixed window length WD or improved B distribution. Inorder to decrease the noise effect, the correlation filter technology is used to process theinstantaneous maximum output amplitude signal of PWD or improve B distribution.Simulation result shows that this method can extract the pseudonoise code under the SNRless than 0dB.(3) The theory of FM-AM transform based on the linear filter output and itsapplication in the extraction of the time-frequency characteristic on pseudonoise code fuzesignal are discussed. Frequency computation methods based on the several ideal types offilter (including filter banks) are presented. The filter banks are designed for thepseudonoise code fuze signal analysis. Simulation results show that this method can extractthe frequency information well and keep the information of phase abrupt point leaded bypseudonoise code modulation. The advantage of this technology is the easy hardwareimplementation because it is not necessary to do Hilbert transform to the signal when thefilter coefficients are in complex form.(4) The technology of inner pulse characteristic extraction and parameter estimationfor pulse pseudonoise code fuze signal based on the methods mentioned in (1) and (3) ispresented. Simulation result shows that these two methods can extract not only the pulse position which can induce the characteristic of PN code but also inner pulse characteristic.The code type analysis method is also discussed.(5) The auto-recognition method of the pseudonoise code fuze signal is discussed. Thesix recognition features are advanced referring to the recognition methods of generalcommunication signals and combining with the specialty of the signals under discussion.The decision machine is used to classify the signals. The simulation results show that themethod can well classify the signals under discussion.(6) The separation and parameter estimation of multicomponent pseudonoise codefuze signals are discussed. Two methods are proposed. One is independent componentanalysis and the other is time-frequency analysis. The former adopts the fixed-pointalgorithm which is fast and stabilization, and then the methods introduced in (1)～(5)could be used to estimate the parameters of the every single separated source. The latteralgorithm is based on the improved B distribution. The simulation results show that theanalysis problem of multicomponent pseudonoise code fuze signals can be solved well bythe two separated methods proposed in this dissertation.