Key Technique Research For HF Burst Communication System Based On BICM-OFDM Modulation

High frequency (HF) communication is an important communication method, especially in military communication, this is attributed to its simpleness, moderate cost, ability to provide near global connectivity without repeaters, beyond line-of-sight, and ease of proliferation. But with the development of modern high technology war, traditional HF communication is threatened by modern electronic countermeasures, so it is very useful to develop efficient anti-jamming measures and anti-interception in HF communication. The object of this dissertation is to study the application of Bit interleave coded modulation (BICM) combined with Orthogonal Frequency Division Multiplexing(OFDM) in HF communication and relevant key techniques based on the model of a new HF adaptive burst communication system (ABCS) proposed by academician Lu Jianxun in China Engineering Academy.The dissertation firstly introduces the new techniques in modern HF communication, and put emphasis on the introduction of ABCS, including its basic theories, main characters and advantages over traditional HF communication system. The application of BICM combined with OFDM in HF adaptive burst communication system was discussed. The following items give the summarization of the main work:1. The modem scheme was realized. The combination of BICM and OFDM is a well known technique to efficiently combat frequency-selective fading channels. The technique of BICM combined with OFDM was applied to HF adaptive burst communication system in this dissertation. The BICM-OFDM system based on hard decision feedback decoding was implemented and the performance was analyzed and simulated, which applied the bases to the choice of modem scheme.2. The receiver must acquire accuracy Channel Status Information (CSI) which could be used to eliminate the effect of frequency-selective fading channel. Based on maximum-likelihood (ML) estimation, two improved algorithms were proposed.1) Using the hard-decision information provided by the decoder of BICM, the information between the channel estimator and the decoder can be exchanged.2) The algorithm which employs low pass filtering in a transform domain was applied to channel estimation. The effect of intercarries interference and additive white Gaussian noise are significantly reduced. The two proposed algorithm efficiently improve the accuracy of channel estimation and get better performance of Bit Error Rate (BER).The conventional superimposed pilot algorithm has drawback of low accuracy. An optimum superimposed pilot which possessing optimum Peak-to-Average Power Ratio (PAPR) was chose. Two improved algorithms were proposed.1) The initial channel estimation employs the first-order statistics. Based on maximum-likelihood, the channel estimation can be acquired iteratively by using the feedback information provided by the decoder of BICM.2) Using the feedback information provided by the equalizer, the channel estimation can be obtained iteratively based on maximum-likelihood. Compared with conventional superimposed pilot algorithm, the two proposed algorithms can improve the accuracy of channel estimation and BER performance.3. Frequency offset could lead to intercarries interference and degrade the performance of system. A new blind frequency offset estimation algorithm was proposed. The proposed algorithm uses the feedback information provided by the decoder of BICM as virtual training sequence. The frequency offset estimation can be acquired by the feedback information. The precision can be improved by the process of iterative decoder. When the iteration number and Signal-to-Noise Ratio (SNR) increase, the frequency offset can be estimated efficiently. Based on maximum-likelihood (ML) estimation, an improved algorithm was proposed. The initial estimation employs the conventional ML algorithm. The remained frequency offset can be obtained by the feedback information provided by the decoder of BICM. The proposed algorithm improves the precision of frequency offset estimation and the range of frequency offset estimation under low SNR.4. A new training symbol was designed, which can be used to obtain the symbol timing estimation and frequency offset. A new frequency offset estimation algorithm was proposed by the characteristic of training symbol whose magnitude is constant. The timing offset can be obtained by using the special structure of training symbol.Based on this special designed training symbol, a new algorithm was proposed which employed differential encoding. Carrier frequency offset estimation was carried out without requiring channel knowledge. By setting a rang of frequency offset, the frequency offset can be acquired by endless searching. The two proposed algorithm need only one training symbol. Compared with conventional algorithms, The proposed algorithms improved the precision of frequency offset estimation and the efficiency of bandwidth usage.