A Study On Blind Despreading Of HF Direct-Sequence Spread-Spectrum Signals

Direct Sequence Spread Spectrum can enhance the anti-jamming performance and data transfer rates of HF communication system, which is widely used in HF communication systems, such as the third generation of HF communication protocol and the US army 141B protocol. However spreading sequence is unknown in advance in non-cooperative scenes, for instance communication reconnaissance, localization tracking of illegal HF communication transmitters and the cognitive radio field. Blind estimation is needed to recover the messages. Therefore, a study on blind despreading of HF DSSS signals is very important.The problem of the blind despreading of HF DSSS signals is studied in this paper. The main idea is to achieve the mathematical optimization model and optimization solution of the blind despreading of HF DSSS signals, based on the characteristics and mathematical models of different HF channels. The researches are as follows:1. In the single channel receiving condition, for the HF DSSS signals propagated by the ground, the channel can be modeled as a Gaussian channel if the channel environment is good. For the HF DSSS signals, the character of symbolic finite sets of spreading sequences is ultilized to solve the blind despreading problem based on the iterative least squares with projection (ILSP) algorithm. The method can be used in different scenes, such as the synchronous short HF DSSS signals, the asynchronous short HF DSSS signals and the synchronous long HF DSSS signals. For the blind despreading of signle-user long HF DSSS signals, a method based on semidefinite relaxation is studied. A joint maximum likelihood estimation model of the spread spectrum sequences and the information sequences is established. And the maximum likelihood estimation problem is transformed into a semidefinite programming problem by relaxing the constraints. Approximate solution is gotten by using the interior point algorithm.2. In the multi-channel receiving condition, for the HF DSSS signals propagated by the ground, ICI exists in the received signals due to the multipath effect, but the ISI between the information symbols can be ignored. For the synchronous short HF DSSS signals, a blind despreading method based on the low rank decomposition is introduced. The synchronous long HF DSSS signals can be modeled as a three order tensor model with missing-data, and is decomposed to low-rank tensors based on the alternating least squares (ALS) algorithm.3. In the multi-channel receiving condition, for the HF DSSS signals propagated by the ionosphere, the ISI is serious due to the large transmission delay, and it can influence some information sequences. For the synchronous short HF DSSS signals where the multipath reflection occurs in the far-field, a Rank_LL1 model is studied. And the method is extended to the asynchronous short HF DSSS signals for the first time. For the synchronous short HF DSSS signals where the multipath reflection occurs in the near-field, a block component model (BCM) model is introduced. And low-rank decomposition can be used for the synchronous short HF DSSS signals blind despreading.