Research On Signal Modulation Recognition Algorithm Under Multipath Channel Condition

Software radio technology and digital band-pass modulation technology have developed very maturely.Scholars and researchers from various parties have fully utilized basic information such as the amplitude,frequency,and phase of electromagnetic waves to derive various modulation technologies.Common modulation methods can be divided into: Amplitude shift keying(ASK),Frequency-shift keying(FSK),phase-shiftkeying(PSK),quadrature amplitude modulation(QAM),etc.In the military and civilian fields,it is significant to be able to successfully identify so many signal modulation methods in the case of non-cooperative communication.As far as the communication environment is concerned,the communication channel environment is becoming more complicated for various reasons while the communication signal style is diversified.In many cases,such as mountainous areas,indoor wireless transmission,mobile communications and other communication environments,the presence of a large number of reflectors and scatterers makes the multipath effect obvious.At this time,due to problems such as signal or channel model mismatches in conventional modulation recognition algorithms,the recognition effect is greatly deteriorated,and even some algorithms cannot achieve the expected recognition effect.Therefore,research on modulation and identification technology that can be applied to communication signals under multipath channel conditions has practical significance in the current communication environment.First of all,we study the modulation identification of single carrier signal in multipath channel conditions.As a pre-processing,the carrier frequency estimation of the signal is realized by using the cyclic spectrum which is less affected by the multipath channel,and the effectiveness of the carrier frequency estimation algorithm is verified by simulation experiments.In terms of signal modulation method recognition,in order to improve the in-class recognition effect of MFSK signals and the inter-class recognition effect of MFSK and other single-carrier signals under multipath channel conditions,we propose a method of extracting feature parameters using the complexity features of the cyclic spectrum section.Compared with the existing methods,the intra-class discrimination of 2FSK 4FSK 8FSK and the inter-class discrimination of MFSK and other large classes can be effectively realized,and the anti-noise performance is better.Aiming at the problem that the conventional inter-class recognition method of 2ASK and MPSK decreases under multipath channel conditions,we have proposed a method using wavelet transform entropy feature to extract feature parameters.Compared with traditional methods,the inter-class recognition effect of 2ASK and MPSK signals has been improved in anti-noise performance.Then we combined the identification features of highorder cumulants and cyclic spectrum extraction,which are also robust to multipath channels,to realize the modulation of nine signals in multipath channel conditions.The recognition rate simulation of the algorithm is performed using a classifier based on threshold decision.The final experimental data shows that all signals can reach 90% recognition accuracy when the signal-to-noise ratio is 4d B.Secondly,we studied the inter-class recognition algorithm of multi-carrier signals and single-carrier signals under the condition of multi-path channels when the identification signal set contains multi-carrier signals—Orthogonal Frequency Division Multiplexing(OFDM)signals and OFDM subcarrier modulation recognition algorithm.In this paper,two algorithms based on high-order cumulants and wavelet transform are used to realize the inter-class recognition of OFDM signals and single-carrier signals.Subsequently,as a pre-processing step of subcarrier identification,the cyclic spectrum and cyclic prefix were used to estimate the parameters of the most critical effective symbol length,chip time width and total symbol length of the OFDM signal.The cumulative quantity and the high-order moment realize the identification of the modulation method of the OFDM subcarrier signal under the condition of the multipath channel,which can complete the distinction between the empty subcarrier and the modulation subcarrier of the OFDM subcarrier and the modulation method of the modulation subcarrier {QPSK 16 QAM 64QAM}.Simulation results show that when the signalto-noise ratio is 2d B,the correct recognition rate of the modulation method can reach 90%.Finally,we realized that when there are three types of multi-carrier OFDM signals {QPSKOFDM,16QAM-OFDM,64QAM-OFDM} and 8 types of single-carrier signals {2ASK,2FSK,4FSK,8FSK,BPSK,QPSK,8PSK,64QAM},single-carrier and multi-carrier inter-class identification of OFDM signals and single-carrier signals,and identification of their respective intra-class modulation methods.Simulation results show that the signal can reach 90% recognition accuracy when the signal-to-noise ratio is 4d B.