Research On Networked Radar Cooperative Detection And Jamming Suppression

A radar is an important means to obtain information in the modern battlefield.The traditional monostatic radar obtains the limited target information due to the single observation angle.In the face of the rapid development of electronic countermeasures technology,the monostatic radar is obviously inadequate in survivability and anti-jamming ability.Networked radars have great advantages in target detection,anti-electronic reconnaissance and antijamming capability compared with the monostatic radar.Based on the networked radar system,the cooperative detection and jamming suppression for networked radars are researched in this thesis,including spatio-temporal alignment,fusion and jamming suppression.The main research and innovative achievements are as follows:1.Spatio-temporal alignment is an important prerequisite for networked radar cooperative detection.The position and attitude angle biases of networked radars will deteriorate the performance of target tracking after spatio-temporal alignment and fusion.For networked radars on static platforms,a spatio-temporal alignment without prior spatial information(including radar positions,attitude angles)and time delay is proposed.The parameters to be estimated include the rotation matrix,translation vector and delay of the networked radars.Based on the target trajectories measured by two radars in the network,the minimum error function in the unified spatio-temporal coordinate system is established.The spatio-temporal registration parameters are estimated by alternate iteration of space and time,and the CramérRao Lower Bound(CRLB)of the corresponding parameters are derived.The statistical errors of the estimated parameters are compared with CRLB.For multi-target scenarios,a multi-target matching and alignment algorithm based on Random Sample Consensus(RANSAC)was proposed.By minimizing the sum of trajectory errors in pairs,the multi-target trajectories of radars are matched,and then all the trajectory pairs satisfying the matching relationship are used to optimize the spatio-temporal registration parameters.The algorithm does not need the positions,attitude angles and the time asynchronous relationship of radar stations.The alignment can be completed only based on the target trajectory measurements,avoiding the errors of Global Positioning System(GPS)and Inertial Measurement Unit(IMU),and reducing the cost.2.For networked radars on moving platforms,the relative positions and attitude angles between the radars are time-varying.A spatio-temporal alignment algorithm based on Discrete Cosine Transform(DCT)for networked radars on moving platforms is proposed.We use a linear representation of DCT basis functions to describe the positions and attitude angles(the yaw,pitch and roll)of moving radars.Both the temporal asynchronocity and time-varying spatial parameters of the moving radars are aligned,which is modeled as a minimization problem with an objective function on the errors between the target trajectories measured by the moving radars,which is solved using a steepest descent method.Furthermore,the CRLB of the alignment parameter is derived for moving radars,which is compared with estimation errors of parameters.Different from the existing methods requiring the prior spatial information(positions and orientations)of moving radars at each sample time,the proposed algorithm completes the spatio-temporal alignment based on the target trajectory measurements from different radar stations only using CPI.Additionally,networked radars can be time asynchronous with their respective CPIs and an unknown delay.3.Aiming at the problem of the large data amount required for spatial alignment in echo domain for networked radars,a spatial alignment in range-Doppler domain with low communication rate is proposed.To reduce the radar-to-fusion-center communication-rate,the method of initial constant false alarm rate detection is used to censor the signals in rangeDoppler domain from local radars.Based on the spatial alignment model for the networked radars in geometry,a maximization problem is formulated.The objective function is the crosscorrelation between the range-Doppler domain signals from different local radars.The optimization problem is solved by a genetic algorithm.After alignment and fusion in rangeDoppler domain,the detection probability of the target is improved.4.Signal-level fusion can be performed after alignment in range-Doppler domain,and the fusion is the key to realize cooperative detection.The target echoes received by different radars in the network is statistically independent,and are generally processed using incoherent fusion.In order to achieve the higher SNR after fusion,a cell-based coherent signal fusion in rangeDoppler domain is proposed for networked radars.The fusion weights are different due to different phase differences when the target signals appear in different range-Doppler cells.We need to calculate the weight for each cell separately.An objective function is established to estimate the fusion weight for each range-Doppler cell using Linearly Constrained Minimum Power Fusion(LCMPF).The weight maintains the signal power of the range-Doppler cell in which the target appears fixed after fusion,and minimizes the power of all cells.The range and velocity of the target in each cell are detected by Constant False Alarm Rate(CFAR)detection on the signal fused by the corresponding weight.Simulation results show that the SNRs after fusion using the proposed algorithm are improved and the targets can be detected via CFAR when a monostatic radar cannot detect the targets due to low SNRs.To verify the effectiveness of the proposed algorithm,a practical test is carried out using the existing hardware system.The SNR after fusion of three radars is improved by 4.3 d B in the hardware test,which is consistent with the simulation result.5.Aiming at the complex jamming environment,the range-Doppler domain data after signal processing for networked radars is used as the training set to carry out jamming situation awareness and classify different jamming types.Deception jamming is a great threat to target detection.For deception jamming,a collaborative anti-jamming algorithm for networked radars based on Stretch processing is proposed.In this algorithm,each local radar sends the trajectory points obtained after signal processing(including the true target and deception jamming)to the fusion center.The fusion center uses the trajectory points obtained by the networked radar to perform the same-source-testing and get the estimation values of the true target distance and velocity,which are sent back to the local radars.According to the corresponding time delay and Doppler frequency,the reference echo is constructed.After Stretch processing,the target and jamming signals are separated in the frequency domain,and the jamming at high frequency is filtered out.Then Stretch inverse processing is carried out to recover the signal containing only the target,which is transmitted to the information fusion center for subsequent signal level fusion.The simulation results show that the proposed anti-jamming algorithm can effectively suppress the jamming in both single and dense deception jamming scenarios.Moreover,the algorithm is still applicable to the maneuvering target and multi-target tracking,and radars can track the true target correctly.The implementation scheme of cooperative anti-jamming algorithm for networked radar based on Stretch processing in hardware circuit is presented.The flow of hardware verification experiment is given on the existing semi-physical simulation platform and digital signal processor,and the anti-jamming performance of the algorithm is verified on the hardware.