Research On Separation Algorithm Of The Time-frequency Overlapped Signal In Single-Channel Based On Certain Prior Information

In the field of communication confrontation,the single-channel signal processing is a key technology,due to its low cost in equipment and good concealment performance in battlefield.However,the existing methods for single-channel signal separation are hardly applied to communication confrontation because of complex mathematical calculation and special application object.Therefore,this thesis focuses on the extrme case of underdetermined problem of the single-channel signal separation.And the writer proposes the separation algorithms based on the certain priori information.Thereinto,the so called “certain priori information” refers to the two different priori conditions for mathematical problem of the single-channel signal separation in the research context of a weak signal and a strong signal overlapping in both time domain and frequency domain,one is given the model of the strong signal,and the other is given the modulation of the strong signal.The main work and innovation points of this thesis are as follows:1.In this thesis,the writer establishes a mathematical model of the single-channel signal separation based on the characterization of multipath channels,which is more in line with the actual situation that it is not jusr adding the weak signal and strong signal together at the receiving end.Furthermore,the writer sets up a standard based on timedomain waveform,spectrum energy and statistical correlation,in order to evaluate the separation effect of the time-frequency overlapped signal.2.Based on the prior information being the model of the theoretical strong signal,the writer proposes a hybrid algorithm combined joint TDOA(Time Difference of Arrival)/ FDOA(Frequency Difference of Arrival)estimation and adaptive filter by using ECA(Extended Correlation Algorithm),in order to separate the weak signal from aliased signals.In the first place,based on cross-correlation of the signal,the writer uses the cross ambiguity function to estimate the difference of both time domain and frequency domain between the aliased signal and the theoretical strong signals,which aims at compensating the theoretical strong signal,in order to obtain the actual strong signal as the reference input of ECA adaptive FIR.And in the second place,the actual strong signal and multipath signals are subtracted from the aliased signals by ECA adaptive FIR,leaving the weak signal.Finally,the validty and feasibility of the algorithm is tested through simulation experiments by using the modulation noise as experimental object.And the result could be applied to the communication receiver with knowing the interference information from the collaboration.3.Based on the prior information being the modulation of the strong signal,the writer proposes a hybrid algorithm combined the signal reconstruction and adaptive filter by using ECA(Extended Correlation Algorithm),in order to separate the weak signal from aliased signals.Firstly,the theory of signal reconstruction is based on the fact that the principal component of aliased signals is the strong signal,in other words,the writer considers the multipath signals and the weak signal as the interference signal so that the aliased signal is capable of being demodulated by using the demodulation method of the strong signal.Meanwhile,the parameters of the aliased signal are also estimated by using the appropriate estimation algorithms of the strong signal.Secondly,the writer reconstructs the signal as the reference input of ECA adaptive FIR by using the baseband coded formats after the demodulation and the estimation of the parameters of thr aliased signal.In the final step,the actual strong signal and multipath signals are subtracted from the aliased signals by ECA adaptive FIR,leaving the weak signal.Specifically,the writer analyzes separately the influence of the power ratio between the strong signal and the weak signal,the comnbination of analog modulation signal(AM / FM),the different combination of digital modulation schemes(ASK / PSK / FSK)or SNR on BER.The result could be applied to the wideband reconnaissance receiver in the field of communication confrontation.