Adaptive Mainlobe And Sidelobe Jammers Suppression Algorithms For Large Scale Phased Array

The extremely huge computational complexity and system cost prevent the realization of element-digitized large scale phased array.Therefore,dimension reduction of large array and rank-reduction adaptive digital beamforming(ADBF)have received considerable attention during the past decades.In practical applications,the ADBF performance suffers from performance deterioration due to the fast relative motion of interferences.Moreover,the sidelobe jammer suppression would have influence on the specially shaped mainlobe for high gain of given area.Even though sidelobe jammers can be rejected effectively by ADBF algorithms,the mainlobe jammer suppression would lead to mainlobe distortion and thus failure of desired signal reception and tracking.Various techniques have been developed for dealing with these above-mentioned issues.However,some essential problems remain to be addressed,which include: i)adaptive performance of subarray configuration approaching the full adaptive performance;ii)broadening and deepening the null in the jammer direction;iii)maintaining the broadened mainlobe while placing null at the jammer's angle;iv)mainlobe maintenance after mainlobe jammer rejection;v)applying adaptive monopulse architecture for mainlobe jammer cancellation to non-rectangular-grid array;vi)adaptive monopulse architecture for mainlobe jammer and sidelobe jammers cancellation and so on.To solve these problems,this dissertation studies on the sidelobe jammer,mainlobe jammer and their composite jammers suppression algorithms for large scale phased array.The main work can be summarized as the following three parts.The first part studies the pattern synthesis algorithms for sidelobe jammers rejection.Firstly,a generalized sidelobe canceller(GSC)weighting approximation at irregular subarrays algorithm is presented for interferences suppression of large scale antenna array after subarray configuration.First the subarray configuration is obtained by particle swarm optimization(PSO).Then digital weights of the second-class subarrays and also the analog weights of the elements and first-class subarrays are optimized in two steps by convex optimization to approximate GSC weights to maximum extent,which are directly applied at element level of the whole array.Compared with direct GSC weighting at subarrays,the proposed algorithm could gain better performance in anti-jamming.Secondly,to broaden and deepen the null in the jammer direction,the covariance matrix tapers(CMT)beamforming algorithm is combined with artificial noise injection method in the ADBF pattern iteratively.That means the covariance of the received data is reconstructed and then adopted in the subsequent adaptive weights computation.Because of the injected noise,the deeper null can be formed at the jammer's angle than the original CMT beamforming algorithm.Therefore,the jammer can be suppressed effectively despite the perturbation of interference location and steering errors.Finally,a novel pattern synthesis method,which broadens the mainlobe and meanwhile forms a deep null in the jammer's direction,is presented.The mainlobe is broadened by iterative FFT(IFT)algorithm.Next,the coefficients obtained in the IFT serves as the quiescent constraint vector in the subsequent linear-constraint-minimum-variance beamforming.The key to the method lies in modifying the constraint vectors by orthogonal projection to eliminate the deep null on the mainlobe maintenance.As a result,the mainlobe preserves broadened and the desired signal is received properly while the sidelobe jammer is rejected.The second part studies the mainlobe maintenance and mainlobe jammer suppression for adaptive monopulse angle estimation.Firstly,a mainlobe maintenance method based on shrinkage estimator is presented to promote the ADBF performance when the mainlobe jammer exists.A blocking matrix is applied to suppress the mainlobe jammer.Then,the linear combination of estimated covariance matrix and identity matrix is optimized to generate more accurate estimation of the covariance matrix.After that,the ameliorative covariance matrix is utilized to generate the adaptive weights to suppress the sidelobe jamming.The proposed method is capable of eliminating mainlobe shifting and high sidelobes caused by blocking matrix preprocessing(BMP).Accordingly,it outperforms other methods in output signal-to-interference-plus-noise ratio,such as noise whitening with error compensation,diagonal loading after BMP and so on.Secondly,an adaptive monopulse architecture for mainlobe jammer cancellation is proposed,which can be applied to triangular-grid array.As we all know,the mainlobe jammer canceller(MLJC)for monopulse estimation can only be used in rectangular-grid array by the virtue of separable property of its product pattern.The linear transformation from triangular-grid array configuration to rectangular-grid array configuration is implemented to solve this problem.Finally,the influences of signal-to-noise ratio,interference-to-noise ratio,direction of arrival(DOA)of desired signal and DOA of mainlobe jammer on the performance of the adaptive monopulse architecture is clarified through theoretical analysis and numerical simulations,which is of great practical significance.The third part proposes two mainlobe jammer and sidelobe jammers suppression algorithms for adaptive monopulse estimation.Firstly,a novel algorithm for suppressing mainlobe and sidelobe interferences based on two sum and three difference channels is proposed.Weighting approximation method is adopted to produce weights to form sum,azimuth difference,elevation difference and double difference beams for irregular subarrays.Then azimuth difference,elevation difference beams and sum beam in the direction of sidelobe interference work as the auxiliary beams of static sum beam,while double difference beam works as the auxiliary beam of azimuth difference and elevation difference beams.Based on the principle of Wiener filter,the auxiliary beams are weighted in the optimized way to cancel the mainlobe and sidelobe interferences in the sum beam and mainlobe interference in azimuth difference and elevation difference beams.The proposed algorithm is of simple structure and high accuracy.It is able to simultaneously suppress the mainlobe and sidelobe interferences without distortion of monopulse ratio.Secondly,an adaptive monopulse architecture,which combines four-channel sum-difference MLJC and improved partially adaptive beamforming technique,is presented in this dissertation in the presence of a mainlobe interference and multiple sidelobe interferences,which is composed of two stages.In the first stage,the blocking matrix is modified to protect the desired signal and mainlobe jammer from being cancelled.The desired overall response which matches the fixed weight response within the mainlobe in GSC adaptation can be achieved.Accordingly,the blocking matrix has no influence on the monopulse ratio.In the second stage,the four-channel MLJC is constructed by employing delta-elevation or delta-azimuth beam as auxiliary beam of sum beam while delta-delta beam as auxiliary beam of delta-azimuth beam or delta-elevation beam.The mainlobe jammer in the sum and difference channels are suppressed simultaneously by adjusting the adaptive weights,which control the auxiliary inputs to minimize the output power of the canceller in a least-squares sense along one direction,while the monopulse ratio maintains undistorted relative to the quiescent monopulse ratio along the other direction due to the separable characteristic of the product weighted array pattern.Therefore,the accurate angle estimation of the target can be yielded by evaluating the adapted monopulse ratio in the presence of a mainlobe jammer and sidelobe jammers.