MAC Protocol Identification And Signal Parameters Estimation For UAV Network

With the development of UAV and communication technology,UAV are widely used in various fields.In order to complete various complex tasks,UAV often appear in the form of swarm.UAV swarm fly within the specified region while there is something wrong in the swarm or the swarm fly into prohibited region,it will bring great potential safety hazards.It is of great significance to conduct reconnaissance of drones.UAV often use frequency hopping signal as flight control signal.In order to prevent potential accidents,it is necessary to receive the frequency hopping signal and extract features from it.Therefore,this thesis studies the identification method of the MAC protocol and the parameter estimation method of the multi-frequency hopping signal.The specific work is as follows.First,taking the Link16 data link as an example,a multi-frequency hopping signal is modeled.Aiming at the blind estimation process of communication reconnaissance,an appropriate time-frequency analysis method is selected to lay the foundation for MAC protocol identification and parameter estimation.Secondly,for the four common MAC protocols,this thesis studies how to use the time-frequency diagram to classify them.In order to select Link16 network from others,it is necessary to classify different MAC protocols.Based on the existing MAC protocol identification method,a method of automatic feature extraction and classification using neural network is designed.The protocol identification under single frequency point and multi-frequency point are discussed.Finally,the parameter estimation and network station sorting of multi-frequency hopping signals are studied.Taking the Link16 data link as an example,the multifrequency hopping signal sent by the swarm is modeled as a synchronous and orthogonal multi-frequency hopping signal.A time-frequency analysis method is used for blind parameter estimation,and parameters such as frequency hopping patterns,the hopping time,and hopping period are estimated.As for synchronous and orthogonal multifrequency hopping signals,a method of network station sorting using DOA estimation is designed.According to the signal characteristics and actual situation,a DOA estimation method using blind source separation in time-frequency single-source is discussed.It can estimate the DOA of the source signal which is larger than the number of receiving antennas.With the use of different antenna arrays,one-dimensional and two-dimensional DOA estimation can be completed.The simulation results show that the above method can be used to estimate the parameters of the synchronous orthogonal multi-frequency hopping.The DOA of the source signal can be estimated under the underdetermined situation.With the use of DOA estimation results,it is necessary to sort different network stations from the synchronous orthogonal multi-frequency hopping signal.