Narrowband Interference Notch Suppression Of Spread Spectrum Communication Systems

From the 50s of last centry,when the military of USA began research about spread spectrum communication,it has been widely applied in military communication,electronic counter,location,admeasurement,mobile communication and so on.Because spread spectrum communication has special advantages such as tenacious ability of resisting interference,information covertness,convenient for encryption and so on.The recon,interference and anti-jamming has become research focus of communication and electronic warfare. The interference technic of spread spectrum communication system is developed energetically in every country.As the mechanism of spread spectrum is to adopt widely frequency band,it applies the possible of interference.It is attacked by invariant narrow-band interference(NBI),time-varying NBI,impulse interfereence and so on.Therefore,the research of NBI suppression in spread spectrum communication has great theoretical and actual meaningness.NBI notch suppression technology of spread spectrum communication comprises time-domain notch suppression and transform-domain notch suppression.Based on the research of people of the past,the object of this paper is issue against the notch suppression technology at present and propose modified methods to improve the performance. As follows it is the main content of this paper:(1)As the adaptive algorithm of time-domain IIR notch filter is sensitive to initial conditions, forgotten factor and smooth factor,a modified method based on DFT frequency estimation is proposed. Received signals with certain length N is transformed by DFT.NBI has banging peak value in frequency domain.According to DFT elements,the gap of two adjacent data is 2π/N . The position k of the max peak value can be fined in frequency domain data with the length 1^N/2.Then the estimator of NBI frequency is obtained.At last it is passed to frequency parameter of IIR notch filter.(2) As the frequency domain notch suppression method based on threshold estimation has bad performance about invariant auto-regresssive(AR) NBI and time-varying NBI, a modified method associated gross threshold and threshold estimation is proposed. Maximum Likelihold (ML) technique isused for estimating N variances of the frequency domain data.Then the variances which higher than gross threshold are filtered out. At last ,the threshold of frequency domain notch filter is achieved by mean value of the variances.(3) In the notch suppression approach based on adaptive wavelet packet transform(WPT),the energy congeries threshold and interference rejection threshold is unknown in reality,and removing the subbands disturbed by interference can lose many useful signals. Therefore a notch suppression approach based on wavelet packet absolute decomposition is proposed.The received signals are absolutely decomposed by wavelet packet,until the deepness arrives at maximum resolution ratio.ML technique is used for estimating variances of the leaves,and the interference rejection threshold T is achieved by minimum variance.If the variance of one leaf is higher than T ,the leaf will be cut back,and the interference will be suppressed. At last,the primitive signals are picked up by inverse wavelet packet transform(IWPT).Simulation results show that the modified methods (1)³ have better performance than former ones.