Research On Measurement And Parameter Estimation Algorithms Of 3D Space-time Wireless Channel

3D MIMO is acquiring increasing attentions in high-speed wireless communication systems. To better depict the singal propagation in 3D MIMO systems, new channel model and its related parameters are needed. Therefore, developing measurement and parameter estimation techniques for 3D space-time wireless channel are of necessity.In this paper, we first take an overview of the wireless channel measurement techniques. Among these approaches, the PN-sequence based channel measurement scheme is the most resprensitive one. As examples, the RUSK MIMO channel sounder and Propsound multi-dimenssional channel sounder are introduced in detail to illustrate the framework of PN-sequence based channel measurement techniques. Moreover, a self-developed MIMO channel sounder is proposed, of which the signal design, hardware structure as well as signal processing procedure are presented specifically.The channel parameter estimation algorithms are also investigated in this paper. We pay our major attentions on the spatial parameter estimation, and some famous algorithms, such as MUSIC, ESPRIT, ML and SAGE, are introduced in both theories and simulating practices. While in real applications of channel measurement, these algorithms are ineligible because of large number of multiple paths, signal correlationship, or the computing complexity. Thereby we propose two new angle estimation algorithms, namely the MERIA algorithm and MAPS algorithm. Both of them are built based the array impulse response(AIR), while MERIA makes use of the rotational invariance of AIR matrix, and MAPS utilizes the orthogonality between the steering vector and the reference subspace, which is obtained from the covariance matrix of AIR vector. Specially designed for channel measurement, MERIA and MAPS can accurately estimate Do As of large number of multi-path signals with relatively low requirements for signal SNR and antenna array size. Detailed theoretical analysis and simulations are presented to verify the performance of proposed algorithms. Additionaly, we implemented MERIA and MAPS algorithms in some testing measurement campaigns, and the results also demonstrate their effectiveness in real applications.