Research On Key Technology Of Non-Cooperative Communication Signal Digital Demodulation

In non-cooperative communications, because the receiver is ignorant of some parameters of the received signal, or even no information of the received signal can be obtained, the accurate synchronism can not be conducted so as to realize the coherent reception. However, we do need to detect the signal and extract the parameters that are required by the demodulation to recover the useful information. Therefore, researches on the digital demodulation technique of non-cooperative communication signals prove to be essential.Digital demodulation techniques of non-cooperative communication signal involve following key parts: the first part is the pre-process of received signal aiming at the elimination of the negative effect caused by multipath propagation and environment noise existing in wireless communication system; the second part is the carrier recovery, including both the compensation of the frequency offset caused by propagation and the track of the phase; the third part is the modulation pattern recognition of received signal to facilitate the demodulation of the desired information.First of all, the principle of LMS adaptive filter is investigated. Considering the simple structure and easy implementation on hardware of LMS adaptive filter, in this thesis, we use it to realize the signal pre-process module. Then, based on the digital carrier recovery method, the frequency offset and phase offset caused by transmission are estimated. For the frequency offset, non-decision data-aided algorithm, frequency offset observation algorithm and M&M algorithm are studied. For the phase offset, instead, the Maximum likelihood estimation (ML), Decision-directed estimation(DD), Power law estimation(PL) and Histogram algorithm (HA) are investigated, including the analysis of their performance and the discussion on possible hardware implementation. In addition, this thesis investigates the characteristic parameter algorithm based on time domain and frequency domain and the high-order cumulant algorithm. The modulation pattern recognition of several modulated signals is discussed here and the results show that the samples of these signals can be classified correctly. In the end, the functional simulation of the NLMS adaptive filter and Decision-directed estimation of phase offset are implementation on FPGA.It can be concluded from the results provided in this thesis that because of the simple structure and low cost of FPGA implementation, the digital demodulation techniques of non-cooperative signals implemented through digital receiver are proved to possess very good application outlook.