Multi-signals Detection In Wideband And Modulation Parameter Analysis Of Mixed Signals In Single Channel

In satellite communication system,the electromagnetic spectrum resources are increasingly strained,and the same communication frequency modulation signal on the more intensive,time frequency aliasing exists.In the broadband receiver under the background of non cooperative reception conditions,single channel receiver is the most common way.Topic for this kind of single channel mixed signal detection and parameter estimation problem,under the condition of non cooperation receiving,from which broadband multi signal detecting,providing for the existence of a judgment,then from the point of view of the cyclostationary analysis,designing algorithm to estimate modulation parameters and determine its modulation mode,realized by the {BPSK,QPSK,OQPSK,8PSK,8qam,16 QAM,16APSK,32 APSK,CPM} mixed single channel mixed signal blind detection and modulation parameters analysis,for the subsequent signal separation,demodulation decoding and reducing information processing lay the foundation for expansion of single channel mixed signal analysis of the breadth and depth of the study.In this paper,the main work and innovation are as follows:1.Aiming at the problem of wideband signals detection,In this paper,a novel spectrum sensing algorithm based on Multi-scale Power spectral density Subband Gradient(MPSG)is proposed.The proposed algorithm,not relying on prior information,calculates gradients of sections for the power spectrum and detects the narrowband signals by adaptive double threshold,which improves the stability of wideband spectrum sensing through multi-scale technique.The paper deduces theoretically expressions of statistical properties,false alarm probability,detection probability and decision threshold in different channels.The theoretical analysis and simulation results prove that: the novel algorithm overcomes effectively the color noise with low computational complexity and strong real-time in Gaussian noise and fading channel,which could accomplish occupying band range localization of prime users.In addition,the proposed algorithm is robust for noise variance uncertainty,even in the low SNR environment.2.Given the concept and mathematical model of single channel mixed signal under the condition of non cooperative receiving.Introducing cyclostationary theory basic concept,the relationship between the statistics and relevant properties,this paper finally analysis common satellite communication signal cyclostationarity,which deduces signal cyclic accumulation and cyclic spectrum of expression in the same,so as to lay a foundation for the following parameter estimation and modulation classification.3.Considering the problem of parameter estimation in single channel mixed signal,this paper presents a fast algorithm for joint estimation of parameters based on cyclic accumulation and improved cyclic spectrum estimation.We propose complete algorithm flow of parameter estimation,which computing high cyclic cumulant and center of power spectrum so as to decide number and carrier frequency of mixed signals,then extracting the peak of symbol rate with proposed cyclic spectrum,finally accomplishing parameter estimation of mixed signals in single channel.Paper proves the statistical features of improved algorithm theoretically.The theoretical analysis and simulation results prove that: The algorithm does not require a priori information such as synchronization code,and suit for Gaussian noise and flat fading channel,even reduce the computer complexity,storage space,estimation variance and data quantity requirement,verifying the feasibility and effectiveness of the algorithm.4.In view of single channel mixed signal modulation recognition problem,under estimated carrier frequency and symbol rate of mixed signal,we use cyclic frequency characteristics of different signals to distinguish different signals,and bulid the signal amplitude templates to match recognition for signals with no obvious cyclic frequency characteristics.Directing at difficult problem of distinguishing QPSK from 16 QAM,this paper proposes a classification algorithm based on fourth order cyclic cumulant based on energy normalization,and deduces principle theoretically.The simulation results show that the proposed algorithm is effective and feasible by different noise,power ratio,overlap ratio and three-signals single channel multi signal modulation classification under the condition of AWGN channel.5.Constructing the mixed SCCPM signal receiving simulation environment platform based on the DVB-RCS2 protocol,and we test the performance of the signals detection and parameter estimation algorithm proposed by this paper.The test results show that carrier deviation than the symbol rate deviation influence signals demodulation accuracy in parameter estimation module,the parameter correct estimation will be of guiding significance of mixed SCCPM signal demodulation.