Study Of Smart Antenna Beam-forming Techniques For LEO To-Ground Communication

The low Earth orbit(LEO)communication satellite is located in low Earth orbit at an altitude between 750km and 1800km.The Earth subtends an angle of about ±55° from nadir at this altitude.The maximum continuous communication time between the ground user and single satellite is less than 10 minutes with ±55° angle changes.The ideal space-borne multi-beam receiving antenna array should have the following characteristics:(1)In order to realize equal sensitivity communication with the users in anywhere of the visible earth,mainlobe gain isoflux is required.(2)For the angle time-varying characteristics of the ground users and the interferences,the array antenna needs to form acquisition beams with interference nulling tracking ability and track beams with high user gain and interference nulling tracking abilities.This paper mainly focuses on the LEO satellite communication system of wide angle coverage with limited numbers of burst users,which needs fast signal acquisition and high gain signal tracking with space-domain anti-jamming capability.Fast quiescent pattern synthesis algorithms,nulling pattern systhesis algorithms with expected mainlobe coverage and nulling tracking ability,adaptive interference nulling algorithms with pattern maintaining and fast user/interference angles estimation algorithms are researched and verified by experimental results.The main works and innovations of this paper are listed as follows:1.The system architecture of smart antenna with adaptive simultaneously multi-pattern on-orbit reconfiguration and adaptive interference suppression is proposed.In this architecture,multi-beam smart antenna system composed of several acquisition beams,covering the visible earth with isoflex and several track beams,steering to the users;The angles of multiple users and possible strong interferences are estimated in real time;All acquisition and track beams dynamically optimize their beam patterns to generate deep nulls tracking interferences in space domain,meanwhile,all track beams also need to steer to the users for high gain of the communication link.2.The existing high performance pattern synthesis methods based on least mean square error(MSE)are not suitable for space-borne real-time computation for the reasons of computation complexity,convergence speed and tunable parameters selection.To solve these problems,two least-squares fast power approximation(FPA)synthesis methods are proposed,called iterative phase update(FPA_IPU)and direct phase update(FPA_DPU)respectively.The weighted factors of different pattern regions in cost function are determined by pattern power approximation principle.The cost function is more reasonable with clearly physical meaning.Then,an iterative nulling tracking iterative method(FPA_INT)is proposed.FPA_INT has simple weight vector iterative formula to suppress dynamic interferences by pattern nulling tracking.The method can maintain mainlobe gain and keep sidelobe level during nulling tracking,while reducing the amount of computation by 2?3 orders of magnitude compared with the exsiting algorithms.3.Iterative transformations between the weight vector and the pattern array factor are often needed in the process of pattern synthesis optimization,which occupy a large amount of computation.To solve this problem,Iterative Fourier transform(IFT)technique is adopted to accelerate the pattern synthesis.IFT technique is extended to the plane array with triangular grid.The problem and its possible solution of IFT technique are discussed when the element spacing is greater than half wavelength.Then,IFT and subspace projection techniques are used in nulling pattern synthesis for tracking the dynamic interferences,called IFT interference subspace projection(IFT_ISP)method.The mainlobe gain is maintained and the dynamic range ratio of weight vector is controlled robustly and precisely in IFT_ISP.4.The direction of arrivals of all interferences must be obtained in advance to accomplish interference suppression in the framework of traditional pattern synthesis.The adaptive array theory is introduced to pattern synthesis for the case of no priori interference angle information.In this way,a fast adaptive interference nulling method with pattern maintaining(AINPM)is proposed under the framework of linear constraint minimum variance(LCMV)and generalized sidelobe canceler(GSC).AINPM can adaptively suppress interferences and keep the antenna mainlobe gain accurately approximated to that of the quiescent pattern.First,the mainlobe subspace constraint is constructed by the dominant eigenvectors of the defined mainlobe covariance matrix.Then,the quadratic constraint of weight vector is relaxed to obtain the closed-form solution with a diagonal loading sample covariance matrix.Finally the diagonal loading factor is precisely estimated to satisfy all constraints.The maintenance of pattern mainlobe and the reduction of the computation complexity can be achieved simultaneously in AINPM method under GSC structure.5.Fast and accurate sum-difference monopulse algorithms are researched under the demands of burst multi-users and interferences angles estimation under wide angle coverage.First,pattern synthesis optimization algorithms for sum and difference with controllable angle error and angle estimation range are proposed to improve the angle estimation speed.The proposed algorithm can effectively maximize the effective angle range at a given angular error,or get optimal error within a given scope of angle range.Step by step monopulse angle estimation method is proposed based on rough and accurate sum and difference patterns synthesized above.The angle estimation time is greatly reduced,while retaining the accuracy as that of the traditional monopulse estimation.Then,a third-order Taylor series based monopulse angle estimation algorithm is proposed to improve the angle estimation accuracy deterioration caused by the errors introduced by the non ideal slopes of the monopulse ratios.2-dimension monopulse angle estimation formula is detailed derived based on the monopulse ratio third-order Taylor series.6.An L band 19-element triangular grid hexagon array smart antenna experimental verification system with circular polarization is designed and constructed.The proposed algorithms are partially realized at this platform.Then,the automatic performance testing environment in anechoic chamber is established and the main index system of space-borne smart antenna is improved.The feasibility,high-efficiency and stability of the proposed algorithms in this paper for multi-beam smart antenna system are verified.They are fast quiescent pattern synthesis algorithms(FPA_IPU and FPA_DPU),nulling pattern systhesis algorithms with expected mainlobe coverage(FPA_INT and IFT_ISP),adaptive interference nulling algorithm with pattern maintaining(AINPM)and the pattern synthesis optimization algorithms for sum and difference patterns for angle rough estimation.The above experimental results lay a solid foundation for the application of smart antenna technique in satellite communication system.