Study On Single-channel Blind Separation Technology Of PCMA Signals

Paired Carrier Multiple Access(PCMA) is a new capacity expansion technique applied in satellite communication system. The two signals can use the same frequency, time slot or code in bidirectional communication. Owing to that, bandwidth utilization of the system is doubled. In a non-cooperative communication, the signal received by third party is a mixture of two signals. The priori information of channel parameters and symbol sequences is unknown, which makes it difficult to achieve the blind demodulation of PCMA signal. To a certain extent, it improves the anti-intercept ability of system.For PCMA signal can be modeled as mixed signal composed of two digital modulation signals, the blind demodulation problem of PCMA signal can be solved by using single channel blind source separation algorithms in anti-collaboration communication. This thesis aims at the characteristics of the PCMA signal, and does a series of detailed research on single channel blind source separation. The main work includes the following aspects:(1)Targeting at the parameter estimation problem of single channel PCMA signal,the thesis introduces a approach to estimate residual carrier frequency and code rate based on second-order cyclic cumulants. The second-order cyclic cumulants is used to estimate the carrier frequency and code rate of PCMA signal.(2)Aiming at the blind separation of PCMA signal in non-cooperative communication, a blind single channel separation algorithm based on independent component analysis(ICA) is studied. Firstly, the parameter estimation is applied to get the code rate and carrier frequency of single PCMA signal, then two baseband mixed signal is obtained by processing. Secondly, the original baseband signal is isolated by using fast independent component analysis algorithm. Finally, the data symbols can be estimated by sampling judgment. The two source information sequences are demodulated from the received PCMA signal without the precondition of information sequence.(3)To separate PCMA signal received in the non-cooperative communication, this thesis studies a blind single channel separation algorithm based on modified particle filter algorithm. The received signal is processed in frame format,then the blind separation is realized in two steps. Firstly, the important weights of particles are computed, and some particles which make a small contribution to posterior distribution in the process of parameter estimation for PCMA signal are abandoned by setting threshold. Thus, the computation complexity of the algorithm can be reduced as the total number of particles changes dynamically. Secondly, the maximum likelihood criterion is used to realize blind separation of two uplink signals after the estimation of parameters. The computation of algorithm can be further reduced by abandoning some particle trajectory with poor quality in the process of calculating the likelihood function of symbols.(4)To solve the blind separation of PCMA signal in non-cooperative communication, this thesis studies a single channel separation algorithm based on modified per-survivor processing(PSP). The initialization part is modified by introducing more precise estimation method of parameters. Then a data-aided estimation technique for unknown parameters is embedded in the structure of viterbi algorithm. Thus, the blind separation of PCMA signal is realized.