Research On UAV Positioning Correction And Routing Under GNSS Interference Environment

Unmanned Aerial Vehicle Ad-hoc Networks(UANET)is an autonomous communication network with flexible networking and strong self-healing capabilities.However,traditional ad-hoc network routing protocols may experience performance degradation when it is applied to UANET due to the high mobility of nodes and frequent changes of network topology.Combining geographic location with routing protocols is a common improvement method,but it requires an accurate position.Therefore,this paper studies and implements methods for GNSS interference suppression in drones,accurate drone positioning,and UANET routing protocol,and the main works are as follows:When facing the interference problem of Global Navigation Satellite System(GNSS)in real environments,this paper proposes an interference suppression algorithm for this problem and applies it to GPS.The algorithm use the Zak transform to convert the GPS signal to the Delay-Doppler(DD)domain,where sub-signals forming the interference signal are solved.Because of the sparsity of the interference signal in the DD domain,the estimated sub-signals of interference signal can be obtained based on the center frequency.The sub-signals are then transformed back to the time domain and subtracted from the GPS received signal to obtain the desired GPS signal.Simulation results show that compared with other interference suppression technologies,this algorithm can effectively improve interference suppression performance even in the presence of high-power interference signals.Furthermore,since the UAV positioning still has some errors after the GNSS interference suppression algorithm is applied,a Reinforcement Learning-based Adaptive Kalman Filter(RLKF)positioning method is proposed to satisfy the requirement for accurate geographical location information in using geographic location-based routing in UANET.This method uses a reinforcement learning model to learn the changes rules of noise covariance matrix.Before updating Kalman filter parameters,the action output of the reinforcement learning model is used to adjust the noise covariance matrix,enabling the Kalman filter to change the noise covariance matrix according to the environment.Compared with the classical adaptive localization method – Adaptive Extended Kalman Filter(AEKF),RLKF achieves higher localization accuracy with reduced maximum localization error,smaller mean and standard deviation.Finally,this paper proposes a Communication Stability Metric Routing(CSMR)protocol based on geographic location information.The protocol considers the communication stability metric of the link between nodes and the distance between the next-hop node and the target node when selecting the next hop.Therefore,the next hop node with the highest stability and closest to the target node is chosen to reduce network transmission delay,improve data packet delivery rate,and network throughput.To verify the effectiveness of the protocol,this paper analyzes it using the Exata network simulation platform and compares it with the Speed and Position Aware Dynamic Routing(SPDR)and the Hybrid Opportunistic and Position-Based Routing(OPBR)protocols.