The Study And Simulation Of The Automatic Recognition Technology Of Aircraft Control Signals

Spacecraft monitoring and control communication is nowadays a hot research incommunication field, especially satellite communication which has been applied in anumber of occasions both in the national defense and civil construction fields. With theprosperous development of satellite communication system, integrated communicationwith different kinds of services has become the main develop direction of satellitecommunication, which will be interconnected with the various ground servicecommunication network. Due to the cross of variety of communication systems, higherrequirements have been put forward to automatic identification of communicationsignals in the monitoring and control communication system. Therefore, modulationrecognition technology is extremely important. Based on the above analysis, this thesisexpands the research in the automatic identification technology of typical formats ofmeasurement and control modulated signals.After in-depth research and comparison of different modulation recognitionmethods, two kinds of automatic identification solutions are designed, aftercomprehensive comparison of the pros and cons, technology based on the theory ofcyclic spectrum correlation was adopted to carry on modulation recognition. This thesiscompletes calculation of spectral correlation function of typical modulation signals toserve as groundwork for the subsequent simulation, and Matlab is used to accomplishsimulation and implementation of FAM algorithm. This thesis finishes the cyclicspectral density function estimation of selected ten kinds of typical modulated signals,displaying them in2D graph and different angles3D graph, echoing the several valuesof feature points from the theoretical calculations and achieving the desired goals.In this thesis, ten kinds of typical spacecraft monitoring and control of themodulation signals, including AM, FM, PM,2ASK,2PSK QPSK, OQPSK,2FSK,MSK,16QAM, are selected to focus on in the research. If purely through themorphological characteristics of the spectral correlation function, these modulationsignals can not be accurately distinguished. In order to solve this problem, and in viewof convenience and practical needs, this thesis firstly redefines a new normalized cycle spectral density function, and based on this concept, the thesis also puts forward threecharacteristic parameters innovatively——γ_x~α（f） δ_x~α（f）and κ_x~α（f） Thesecharacteristic parameters constitute a feature set, and this set is used to designreasonable modulation recognition process, realizing the accurately recognization of theselected ten kinds of signals. The thesis uses hybrid programming of Matlab andLabwindows/CVI to verify the feasibility of the new feature parameters and newmodulation recognition methods. Experiments show that, the proposed method canreach identification precision of98%under the condition of SNR=10dB, while the noiseis narrowband white gaussian noise. That’s to say, the characteristic parameters for theautomatic identification of the signal are feasible, and is easy to be implemented.Therefore, expected goals are achieved.This thesis do some useful attempts and exploring in modulation recognitiontechnology, research results of this thesis can not only serve the spacecraft, to expandthe communications business in the area of signal processing, the results also have avery broad prospects for the applications of unmanned stations.