Research On Channel Modeling And Signal Modulation Recognition For UAV Reconnaissance Based On Alpha Stable Distribution Theory

The unmanned aerial vehicle(UAV) has been increasingly widespread used in civilian frequency monitoring and military electronic countermeasure as a new type of equipment. However, because of the complex electromagnetic environment and non-cooperative position the UAV work in, there will be serious pulse noise and multipath fading in its reconnaissance channel, which lead to two problems: on one hand, traditional model cannot accurately describe the UAV reconnaissance channel; on the other hand, pulse characteristic of the noise in the reconnaissance channel results in the existing signal processing method based on Gaussian distribution invaliding. Aiming at these problems, Alpha stable distribution theory was introduced to research on the Ultrashortwave and HF channel modeling and signal modulation recognition for UAV reconnaissance in this thesis. The research results improved existing models and algorithms, which is of important theoretical significance and practical value in non-Gaussian signal processing and UAV application.The innovative achievements obtained in this thesis can be briefly summarized as follows:1. According to the fading mechanism of Ultrashortwave wireless channel, aiming at the impact of the complex electromagnetic environment and non-cooperative position to UAV reconnaissance channel, traditional wide-sense stationary-uncorrelated scattering(WSSUS) model was corrected to model UAV Ultrashortwave reconnaissance channel. Alpha stable distribution noise was used to simulate the noise environment, which improved flexibility and practicability; and WSSUS system parameter determination based on flight conditions and environment was introduced to make the model conform better to the practice. Monte Carlo simulation proved correctness and effectiveness of this model.2. According to the characteristic of HF ionospheric channels and the properties of UAV movement, improved HF broadband ITS model was adopted to model UAV HF sky-wave reconnaissance channel,and some suggestion on UAV practical application was proposed in accordance with the characteristic of UAV reconnaissance channel. The novel model overcomed the faults of tranditional models in flexibility, complexity and accuracy: Alpha stable distribution was applied to simulate noise environment and random modulation function, which improved flexibility; the method based on anchor-poinits fitting was applied to simulate time delay power distribution, which reduced complexity; and definite phase function was improved on the basis of UAV flight condition, which increased accuracy. Simulation based on measured data proved the performance superiority of this model.3. To solve the problem that traditional second-order statistics invalid in the condition of Alpha stable distribution noise environment existing in UAV reconnaissance channel, the concept of real normalized compression function(RNCF) was proposed, and the properties of generalized second-order cyclic statistics based on RNCF were also deduced, on the basis of which the amplitudes of generalized second-order cyclic spectrum in specific frequencies and cyclic frequencies were extracted as the feature parameters, and the minimum error criterion was used as a classification algorithm to achieve modulation recognition. Simulation results show that this method has good performance in both Alpha stable distribution noise and Gaussian noise environment; moreover, it has similar performance as other methods based on generalized second-order cyclic statistics using non-linear transform, but has lower complexity, stronger robustness, and higher flexibility.4. Multi-path effect of UAV reconnaissance channel was further taken into account on the premise of Alpha stable distribution noise environment; and a new modulation recognition method was proposed, by introducing normalized compression function(NCF) to generate a series of novel generalized cyclic statistics. The method extracted higher cyclostationary feature of the signals by generalized cyclic statistics as decision criterion to achieve modulation recognition in Alpha stable distribution noise environment. Simulation results show that the proposed method has good performance in different channels with single or multiple propagation paths in Alpha stable distribution noise environment. Moreover, it exhibits higher flexibility, better performance but lower complexity than other methods based on generalized cyclic statistics using non-linear transform.Besides,an improved method to produce Alpha stable distribution sequence was proposed based on the properties of Alpha stable distribution in this thesis, which reduced the complexity on the premise of accuracy.