Research On Space Direction Finding For OFDM Signals

Orthogonal Frequency Division Multiplexing(OFDM)is an efficient multi-carrier modulation technique.Because of its resistance against frequency selective fading resistance,high spectrum efficiency and other characteristics.OFDM has been studied by researchers in the field of array signal processing.Effectively estimating the direction-of-arrival of(DOA)of OFDM signal has become a new research direction in the field of array signal processing.Considering the demand of practical application,this thesis mainly focuses on the DOA estimation of OFDM signal in non-uniform noise environment and underdetermined conditions,the main research contents are given as follows:1.The thesis firstly studies some existing wideband direction finding algorithms.Such as Incoherent Signal-Subspace Method(ISM),Modified ISM algorithm and Coherent Signal-Subspace Method(CSM).It is known that these algorithms only applies to the cases when the number of sources is smaller that of sensors.And these algorithms will more or less produce a significant performance deterioration in non-uniform noise environment.Which may provide the theoretical foundation and reference basis for research on estimating the DOA of OFDM signal.In non-uniform noise environment these algorithms suffer from significant performance deterioration.Studies on these conventional DOA estimation algorithms provide theoretical foundation and reference basis for exploiting new DOA estimation algorithms of OFDM signal.2.In the presence of non-uniform noise,some classic direction finding algorithms may yield a series of problems,Such as DOA estimation errors increases,algorithm of angle resolution deterioration,pseudo peak phenomenon and so on.In this thesis,a direction finding algorithm for OFDM signal based on the MUSIC-Like method is proposed.Firstly,virtual array transformation is used.Secondly,the optimal weight vector is obtained by a new beamforming method.Thirdly,the spatial spectrum of OFDM signals is obtained using the weighted average of sub-band spatial spectrum.The simulation results show that the proposed not only obtains more precise estimation in the non-uniform environment,but also is able to estimate DOA of coherent signals.In addition,the proposed can realize direction finding without knowing the sources number,and is not dependent on the subspace decomposition.It can avoid the bad impact of incorrect source number estimation on estimation precision.3.For underdetermined conditions,i.e.,the number of sources is greater than the number of sensors,most of the existing algorithms(ISM and so on)may be unable to estimate the DOA of signals.In this thesis,a direction finding algorithm for OFDM signals based on sparse representation and Khatri-Rao subspace is presented.This algorithm firstly construct an improved nested array to improve the degree of freedom of array;Then,combined with Khatri-Rao product a new virtual array is constructed;Finally,direction finding is realized through sparse reconstruction algorithm.Computer simulation results show that the proposed algorithm yields better DOA estimate results than the conventional DOA estimation algorithms.It also improves estimation precision under underdetermined conditions and in the background of non-uniform noise.