Research On Broadband Multiple Signals Processing And Automatic Modulation Classification

Due to its high flexibility and high reusability,the wideband digital signal monitoring platform based on software radio architecture has been widely used in civil and military fields such as radio frequency regulation,electronic warfare and so on,so it has important theoretical value and practical significance to the research of its related technology.In this paper,the multi signal detection technology,the channel filtering technology,the automatic modulation recognition technology of narrow band signal are deeply studied.The work of this paper is a part of the research project undertaken by the author.The main work and innovation of this paper are as follows:1.This paper introduces the basic theory of software radio reception,including the sampling theory of radio frequency direct bandpass,the theory of random sampling rate change and its efficient implementation structure,digital down conversion theory and its realization structure provides a theoretical basis for the realization and application of the broadband signal monitoring platform based on software radio architecture.2.In view of the problem that the detection threshold is difficult to be determined and the less protective interval is difficult to distinguish,a top-down algorithm for wideband multi signal detection and location based on region growing method is proposed.Regard the wideband signal power spectrum after noise base flattening as a two-dimensional gray image,implant one or more seeds into the signal area,then use the region growing method in the digital image segmentation to determine the different signal region and the noise region.The simulation results show that the new method can separate the signals with small protection interval and is more accurate for the estimation of the number,bandwidth and center frequency of the signal.3.According to the problem of multi-channel signal filter when the number of sub bands,bandwidth and central frequency are changed arbitrarily,an efficient non uniform channel filtering architecture without blind area is proposed.Firstly,the computational complexity of the uniform channel filtering structure based on weighted overlap add(WOLA)method is analyzed,which is suitable for the combination of adjacent signal.Then,the design requirements of the prototype filter are analyzed when the adjacent channel is merged to meet the phase distortion and amplitude distortion.Finally,based on the structure of WOAL uniform channel filtering,a non uniform channel filtering structure is proposed based on the adjacent channel.The simulation results show that the proposed structure can realize multi-channel signal extraction,although the number of sub bands,bandwidth,center frequency are constantly changing.The output flexibility and the reset of the system are very strong.The proposed structure calculation complexity is also very low,suitable for engineering applications.4.For the modulation signal set{BPSK,QPSK,8PSK,8QAM,16 QAM,32QAM,16 APSK,32APSK},the existing identification methods need accurate estimation of the frequency offset.However,the accurate estimation of the high order MQAM and MAPSK modulation signal frequency is difficult.To solve this problem,this paper proposes a new method of anti frequency offset modulation recognition based on Gauss mixture model.In the presence of frequency offset,the probability distribution of the normalized amplitude of the signal symbol sequence is analyzed to meet the Gauss mixture model and the parameters of the mixed model are only related to the modulation of the signal.Finally,the modulation recognition problem is converted to the parameters of the Gauss mixture model.The simulation results show that the new method is more tolerant to the frequency offset and noise,and it is not necessary to set the threshold value in the process of recognition,and the dependence on data length is low.5.The article has completed the digital down conversion module,the channel filter module and the modulation recognition module On the basis of the wideband signal detection platform.Then,function test was carried out to the whole system.Test results show that the digital down conversion module,channel filter module completed the sampling rate conversion,spectrum shifting and channelized filtering according to the requirement of parameter setting and the modulation recognition module completes the stability identification of the signal modulation mode,the intended purpose is achieved.