| The accurate estimation of the arrival angles of receiving signals on the base station side is one of the important prerequisites for improving the location of the indoor terminals.In this thesis,spatial spectrum estimation techniques are used to solve accurate estimation problems of the complex indoor multipath coherent signals and reduce the complexity of the spatial spectrum estimation algorithms.This thesis first introduces the design of antenna array pattern,and gives two classic spatial spectrum estimation super-resolution algorithms: MUSIC algorithm and ESPRIT algorithm.Considering the high correlation or coherence of indoor signals,MUSIC algorithm based on space averaging or smoothing technology is studied,and the complexity of the algorithm is analyzed in detail.Secondly,in order to solve the problem of array aperture reduction and the high complexity of spectral peak search in MUSIC algorithm,the thesis studies the unitary transformation ESPRIT(Unitary-ESPRIT)algorithm,which reduces the cost of computing.This thesis further studies the spatial spectrum estimation technologies with better spectral resolution performance: improved algorithms of EPUMA or RISR algorithm.It is proposed that the combination of EPUMA algorithm and spatial smoothing technology can greatly improve spectral estimation performance of coherent signals.The noise power estimation method of RISR algorithm is given,which preserves the advantages of the algorithm in lack of source numbers.Finally,the maximum likelihood(ML)algorithms with optimal performance and their implementation are studied,i.e.expectation maximization(EM)algorithm and its derivative SAGE algorithm.EM algorithm simplifies the multidimensional search of ML algorithm to one-dimensional search problem,and reduces the complexity.The SAGE algorithm is used to measure the angle information of the signal.This method improves the convergence of the EM algorithm and further reduces the complexity. |
PDF Full Text Request