Modulation Identification And Parameter Estimation For Shortwave OFDM Signals

Due to its small volume,flexibility,low cost and anti-destroy ability,shortwave communication is irreplaceable in modern communication,especially in the field of mi litary communication.Orthogonal Frequency Division Multiplexing(OFDM)can overco me original disadvantages in shortwave communication due to its high spectrum efficie ncy,strong anti multipath ability.So OFDM has became the main modulation in wide band shortwave communication.Shortwave OFDM is widely used in military communi cation,so the study of non-cooperative shortwave OFDM signals detection is of impor tant significance.The non-cooperative shortwave OFDM signals are the detect objects in this paper.The aim is to improve the reconnaissance efficiency of the system.The reconnaissance techniques suitable for the non-cooperative shortwave OFDM signals are proposed and demonstrated.The main research contents of non-cooperative shortwave OFDM reconnaissance system are: OFDM existence detection,parameter estimation and sub-carrier modulation recognition.First step is the detection of OFDM signals.Signal detection is mainly to identify OFDM signal and single carrier signal.By time-domain statistics to recognize the OFDM signals from single carrier,higher order statistics is the feature vector.The joint estimation of higher-order moments and the combination of higher order cumulants of detection in OFDM are compared in rayleigh channel.It is verified by simulation that high order cumulant combination as the detection algorithm is not affected by the modulation of subcarrier and the number of subcarrier.The high order cumulant method is extended to detection of OFDM signals in the shortwave channel.The relationship between feature vector value range and the parameters of the HF channel parameters is derived in common shortwave channel and bad shortwave channel.Through simulation,the validity of the identification method to detect shortwave OFDM is verified.Secondly,the parameters of OFDM are estimated.The main parameters in OFDM demodulation system are: cyclic prefix length,effective symbol length,chip width,carrier frequency etc.The effective symbol length and the symbol cycle length are estimated based on the auto-correlation of OFDM symbols,and then the CP length is obtained.The cyclostationarity of signals is introduced,then the high order cumulant properties of OFDM signal is proved.Chip width and carrier frequency are estimated by cyclic correlation function and cyclic spectrum of OFDM signal.Estimation in the parameter cycle domain has strong anti noise ability.Finally,the OFDM sub-carrier modulation recognition is completed.The sub-carrier recognition model is built after the parameter estimation.The sub-carrier modulation is recognited by different combinations of high-order cumulant.Combined with the reduction of single carrier recognition thought,sub-carrier modulation recognition for high order MPSK and MQAM are proposed,the first step is to reduce the order.