Research On Context Sensing Method Of Vehicles Using Radar And Communication Integration Frameworks

With the rapid development of the national economy,vehicles have become an important part of the modern city.However,traffic congestion,accidents and environmental pollution are becoming social issues need to be addressed.Intelligent driving,or even totally autonomous vehicles are showing to be promising solutions to reduce the probability of accidents and improve road utilization efficiency,which has attracted great attention from both academic and industrial areas.Real-time and accurate information of of both vehicles and surrounding environment(referred to “context information”),is of great importance for intelligent driving applications.Hence,accurate detection and sensing capabilities are required as well as wireless communications for data transmission and fusion.Therefore,the large number of on-board sensors lead to many problems,such as an increase in system size and energy consumption,serious electromagnetic interference.Recently,radar-communication(RadCom)integrated systems have been widely investigated,due to its can reduce hardware cost and save spectrum resources,which is becoming a hot issue of current research.RadCom can realize the versatility,miniaturization and multi-function of vehicle-mounted equipment,and have important theoretical significance and practical applied foreground.This thesis aims at the context information sensing problem for intelligent driving,adopting the high-precision vehicle-mounted radar,according to the needs of high accuracy context sensing and information fusion,we propose an integrated waveform design using the orthogonal frequency division multiplexing(OFDM)scheme.Efficient fusion methods are also proposed,to achieve real-time,accuracy and robustness of the context sensing.In detail,the main contributions of this thesis can be summarized as follows.Phase coded(PC)OFDM waveforms are designed for vehicle applications.Specifically,the signal characteristics are further studied from ambiguity function(AF)and peak-to-mean envelope power ratio(PMEPR).The expression of AF is theoretically derived,and the effect of signal parameters on range AF and velocity AF is analyzed,which formulates a theoretical basis for the integrated signal design.In addition,an improved tone reservation(TR)method is proposed to reduce PMEPR of the integrated signals,where the complementary set on the peak reduction tones(PRTs)is constructed to make AF and PMEPR achieve better performance at the same time.Aiming at the estimation of target parameters for RadCom system in the single vehicle sensing,a low complexity signal processing algorithm with high reliability is proposed.For integrated signals using continuous waveforms,a modulation symbol-based processing algorithm is used,combing with frequency domain oversampling technology,to achieve the joint estimation of the range,velocity,angle and scattering type simultaneously.For integrated signals using pulse train PC-OFDM,a correlation and discrete Fourier transformation(DFT)processing algorithm is proposed,where the phase codes with good autocorrelation characteristics is utilized to reduce the effect of carried information on radar performance.Both high velocity resolution and unambiguity can be achieved using pulse coherent accumulation and LS-based ambiguity removal method.High-speed vehicles result in obvious Doppler,which cause inter-carrier interference(ICI).Besides,the received signals suffer from interferences from other involved vehicles.Aiming at the ICI problem,continuous waveforms are improved,and interleaved OFDM(I-OFDM)integrated signals are designed,and an ICI-free processing for radar is given using Doppler estimation and correction.Furthermore,the mixed signal separation algorithm using signal reconstruction is proposed,to recover the target echos by cancelling the reconstructed signals.Simulations show that,comparing the existed algorithms,the proposed methods can realize an ICI-free processing for radar,better separation accuracy,and higher resolution range-velocity image under low signal-to-interference ratio(SIR).Considering that the vehicle may not achieve enough coverage frequently,we introduce unmanned aerial vehicles(UAVs)assisted implicit cooperative sensing strategy.To detect the weak targets UAVs under low signal-to-noise ratio(SNR),continuous waveforms are improved,and repeated symbols OFDM(RS-OFDM)integrated signals are designed,and a combined Doppler compensation and compressed sensing algorithm is proposed,in which a sparse representation model for the one-dimensional range profile is constructed.In addition,for the data fusion problem of RadCom/GPS/INS observation,the implicit cooperative positioning(ICP)algorithm using dynamic non-parametric belief propagation(DNBP)is adopted,and the proposed algorithm is validated based on Cramer-Rao lower bound(CRLB)in the improvement of performance and complexity.Through this research,it can improve the vehicle location accuracy in the coverage hole.