New techniques for orientational beamforming systems

• Main Author: Han, Jiangyan
• Other Authors: Ng Boon Poh
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Electrical and electronic engineering::Wireless communication systems
• Online Access: https://hdl.handle.net/10356/170082
• Institution: Nanyang Technological University

Space division multiple access (SDMA) is widely used in modern wireless communication systems to substantially increase spectral efficiency and attainable channel capacity. The SDMA technique implements an antenna array to separate different users according to their directions through beamforming. Adaptive beamforming techniques and direction of arrival (DOA) estimation methods have been extensively investigated and adopted in SDMA to adapt the array beam to each user while nulling out multiuser interference (MUI), thereby increasing the channel capacity. Recently, another spatial dimension, named array orientation, has been explored, which provides an additional degree of freedom for further increasing channel capacity and spectral efficiency of SDMA. The orientational beamforming (OBF) system is employed to exploit the new spatial dimension, which spatially filters signals according to the orientations of the transmitting and receiving arrays regardless of the signals’ directions. Therefore, it can effectively separate co-channel users lying in close directions, which, together with low co-channel interference in the array orientation dimension of the spatial domain, further increase the attainable channel capacity. However, essential techniques required for the OBF system to achieve the expected increase in channel capacity have remained to be investigated in-depth so far. This thesis mainly focuses on new techniques for the OBF system to fill the gap. Considering the coexistence of the OBF system and other wireless communication systems, there are two kinds of interference that primarily deteriorate the performance of the OBF system, which are directional interference and MUI. Therefore, this thesis first studies algorithms for suppressing these two kinds of interference in the OBF system. For directional interference suppression, three adaptive OBF algorithms, namely the orientational maximum signal-to-interference-plus-noise ratio (O-MSINR) and orientational linearly constrained minimum variance (O-LCMV) algorithms, and the orientational generalized sidelobe canceller (O-GSC), are proposed by extending adaptive techniques in conventional directional beamforming (DBF) systems (directional antenna arrays) to the OBF system, taking into account the similarities and differences between the OBF system and the conventional DBF system. Since they are constructed on the array orientation dimension of the spatial domain, and orientational constraints instead of directional constraints are used to guarantee an orientational gain, they are independent of the signal direction. This property enables the proposed adaptive OBF algorithms to suppress interference in the same direction as the desired signal, which cannot be easily achieved by current adaptive DBF algorithms. Computer simulations show the effectiveness of the proposed adaptive OBF algorithms in reducing bit error rate (BER) in many narrowband and wideband interference situations. As for MUI suppression, a nonlinear OBF system constructed on the complex correlation-based radial basis function (RBF) neural network, referred to as the RBF-OBF system, is developed. The proposed complex correlation-based RBF neural network evaluates the similarity between an input vector and the center vector of a hidden layer neuron by the signed complex correlation coefficient rather than the Euclidean distance, which is more suitable for the case where the training samples are noiseless in the frequency domain. In the RBF-OBF system, the newly defined noiseless orientational steering vectors are used to train the neural network, contributing to the robustness of the proposed RBF-OBF system to pulse waveform variations. The proposed RBF-OBF system is firstly designed in the additive white Gaussian noise channel, and then generalized to the line-of-sight (LOS) multipath channel. Numerous computer simulations show that the MUI is effectively suppressed using the proposed RBF-OBF system, and the BER of the proposed system is significantly lower than that of the conventional linear OBF system. In the proposed RBF-OBF system for MUI suppression, the DOAs and array orientations of different users are required to be known a priori to train the neural network. In addition, in the OBF system and proposed adaptive OBF algorithms, the enhancement of the desired signal at the receiving array requires the knowledge of the transmitting array orientation of the desired user. Therefore, DOA and orientation of array (OOA) estimation methods are then investigated in this thesis. As indicated by the signal model, the DOA information in the OBF system can be readily estimated using the existing DOA estimation methods applicable to arbitrary arrays, so the thesis focuses on the OOA estimation problem. Several OOA estimation methods based on the precise DOA information are proposed, namely the power spectral density-based least squares method, electronical rotation-based OBF method, and single-source-array OBF method. In each method, the OOAs of different transmitting arrays are estimated individually by searching in the one-dimensional parameter space, instead of jointly estimated in the multidimensional space. Then, the effects of the DOA estimation error on the proposed OOA estimation methods are analyzed. A random search method is proposed to correct the DOA error and realize satisfactory OOA estimation. In view of the fact that despite the requirement of a pair of identical transmitting and receiving arrays, the OBF system has a similar configuration to the conventional multiple-input multiple-output (MIMO) system, we finally compare it with MIMO systems. The major differences can be summarized in two aspects: 1) MIMO systems usually adopt differential space-time modulation and detection schemes to avoid tedious channel estimation, while the OBF system has the intrinsic property of working well without channel state information in the LOS environment due to the matched antenna positions in the transmitting and receiving arrays. 2) The OBF system outperforms the differential detection schemes for MIMO systems in the LOS environment under low signal-to-noise ratios (SNRs). However, the OBF system is rarely effective in the non-LOS (NLOS) environment. To address these issues, a hybrid differential detection (HDD) scheme is proposed by combining the OBF system with a proposed differential OBF (DOBF) system. In the NLOS environment, the HDD scheme is reduced to the proposed DOBF system with a fully differential detection scheme, which can achieve the same performance as the MIMO system based on differential space-time block coding, but with much lower computational complexity. In the LOS environment, the proposed HDD scheme can achieve overall better performance than both the OBF and DOBF systems, provided that the optimal and suboptimal parameters given in this thesis are used to obtain the combination coefficient.