| Underwater acoustic communication has become the research focus owing to the economic and military significance of ocean. The parameters of underwater acoustic channel need to be estimated in order to ensure the reliability and effectiveness of underwater acoustic communication. The data-aided algorithms are adopted in this dissertation, and nowadays existing preambles in underwater telemetry are almost exclusively based on LFM signals, also known as Chirp signals. Therefore, it is significant to research the multi-parameter estimation algorithms of underwater acoustic channel.The in-depth research in this dissertation is based on FRFT algorithms. We treat Chirp signals as the underwater detection signal, and the parameters of underwater acoustic channel are estimated aiming at the two models of received Chirp signals (the constant k model and the variable k model) and according to the relation of Chirp signals and FRFT. The work of this dissertation is as follows.Firstly, three novel algorithms for the multi-parameter estimation of underwater acoustic channel are proposed for the constant k model of Chirp signals. (1) The FRFT based estimation algorithm:The advantage is that it can estimate the precise multi-path number, the delays and accurate Doppler shift. However, this algorithm is just suitable to the situation where each of the paths has the same Doppler shift and the exact delay estimation of direct path is known in anticipation. (2) The double-chirp signal estimation algorithm:The strength is that it can directly determine the multi-path number, the Doppler shifts and the delays regardless of the Doppler shifts being the same, similar or very different. However, the shortcoming is that the estimation errors of Doppler shift and delay are quite large due to the introduction of the fast algorithm of FRFT and it makes computational overload. (3) The joint estimation algorithm:The virtue is that it can directly and accurately determine the multi-path number and the delays and Doppler shifts regardless of the Doppler shifts being the same, similar or very different. And it can also directly and accurately estimate the multi-parameter of underwater acoustic channel no matter how short or long the delays of the received signal are. However, the shortcoming is that the calculation errors need to be searched finely in small range, which results in complex algorithm and computational overload.Secondly, for the variable k model of the Chirp signal, a novel multi-parameter estimation algorithm for underwater acoustic channel based on the Chirp-based character of FRFT is proposed, and then two designs of packet structure based on communication system of FRFT for different waters are also given for the variable k model of the Chirp signal. (1) The Chirp-based algorithm of FRFT:the merit is that we can accurately and quickly estimate the parameters of underwater acoustic channel. The algorithm is high accurate and simple, as well as the speed of calculation is very fast. What is the most important is that it can be implemented in real-time processing. However, the estimation performance will decrease when the Doppler shifts are approximate, and the weak path of the two paths with approximate Doppler shift will be severely influenced, but the problem can be solved by high resolution and long sine signal. (2) The structure design of the first packet suitable for waters of fast change:The strongpoint is that it makes the multi-parameter estimation per package possible, which greatly increases the reliability of data transmission. However, it reduces transmission efficiency. (3) The structure design of the second packet suitable for waters of slow change:The strength is that it has higher transmission efficiency. However, it greatly reduces the reliability.Finally, a lot of simulation examples are given to verify the effectiveness of the estimation algorithms. And the simulation shows extensively how the two packets are processed. Underwater acoustic communication has become the research focus owing to the economic and military significance of ocean. The parameters of underwater acoustic channel need to be estimated in order to ensure the reliability and effectiveness of underwater acoustic communication. The data-aided algorithms are adopted in this dissertation, and nowadays existing preambles in underwater telemetry are almost exclusively based on LFM signals, also known as Chirp signals. Therefore, it is significant to research the multi-parameter estimation algorithms of underwater acoustic channel.The in-depth research in this dissertation is based on FRFT algorithms. We treat Chirp signals as the underwater detection signal, and the parameters of underwater acoustic channel are estimated aiming at the two models of received Chirp signals (the constant k model and the variable k model) and according to the relation of Chirp signals and FRFT. The work of this dissertation is as follows.Firstly, three novel algorithms for the multi-parameter estimation of underwater acoustic channel are proposed for the constant k model of Chirp signals. (1) The FRFT based estimation algorithm:The advantage is that it can estimate the precise multi-path number, the delays and accurate Doppler shift. However, this algorithm is just suitable to the situation where each of the paths has the same Doppler shift and the exact delay estimation of direct path is known in anticipation. (2) The double-chirp signal estimation algorithm:The strength is that it can directly determine the multi-path number, the Doppler shifts and the delays regardless of the Doppler shifts being the same, similar or very different. However, the shortcoming is that the estimation errors of Doppler shift and delay are quite large due to the introduction of the fast algorithm of FRFT and it makes computational overload. (3) The joint estimation algorithm:The virtue is that it can directly and accurately determine the multi-path number and the delays and Doppler shifts regardless of the Doppler shifts being the same, similar or very different. And it can also directly and accurately estimate the multi-parameter of underwater acoustic channel no matter how short or long the delays of the received signal are. However, the shortcoming is that the calculation errors need to be searched finely in small range, which results in complex algorithm and computational overload.Secondly, for the variable k model of the Chirp signal, a novel multi-parameter estimation algorithm for underwater acoustic channel based on the Chirp-based character of FRFT is proposed, and then two designs of packet structure based on communication system of FRFT for different waters are also given for the variable k model of the Chirp signal. (1) The Chirp-based algorithm of FRFT:the merit is that we can accurately and quickly estimate the parameters of underwater acoustic channel. The algorithm is high accurate and simple, as well as the speed of calculation is very fast. What is the most important is that it can be implemented in real-time processing. However, the estimation performance will decrease when the Doppler shifts are approximate, and the weak path of the two paths with approximate Doppler shift will be severely influenced, but the problem can be solved by high resolution and long sine signal. (2) The structure design of the first packet suitable for waters of fast change:The strongpoint is that it makes the multi-parameter estimation per package possible, which greatly increases the reliability of data transmission. However, it reduces transmission efficiency. (3) The structure design of the second packet suitable for waters of slow change:The strength is that it has higher transmission efficiency. However, it greatly reduces the reliability.Finally, a lot of simulation examples are given to verify the effectiveness of the estimation algorithms. And the simulation shows extensively how the two packets are processed. |
PDF Full Text Request