Research On Adaptive Multiuser Detection Technology For MC-CDMA System

The capacity of the Multicarrier CDMA (MC-CDMA) system is limited by the effects of the multiple access interference (MAI). The multiuser detection is a powerful technique to combat MAI and near-far resistant, and reduces the requirement for precision of controlling power. Thus spectral resources can be utilized more effectively, which improves the communication capacity substantially. Adaptive multiuser detection has been receiving a great deal of research recently because of its real-time characteristic.This paper introduces the research situation and the basic princile of multiuser detection, studies the subspace-based minimum-mean-square-error (MMSE) multiuser detection, the adaptive minimum-output-energy (MOE) multiuser detection and the multiuser detection based on constant modulus algorithm (CMA). For such practical problems as slow convergence, bad anti-jamming, low resolution and high computational complexity, some new algorithms are proposed to solve them. Simulation results show the effectiveness of the proposed algorithms.The performance of the MOE multiuser detection will degrade when it is affected by the noise subspace in practical situations. To solve this problem, a noise suppression MOE multiuser detection algorithm is proposed. We define a new cost function based on the MOE criterion, make orthogonal processing between the weight vector and the noise subspace, and use the noise subspace to impose multiple constraints on the new defined MOE cost function. A noise suppression linear conjugate detector is designed, and the recursive least-square (RLS) algorithm is used to obtain the weight vector adaptively. The proposed algorithm mitigates the noise in the weight vector, and exploits the information contained in the pseudoautocorrelation of the observables. Therefore the performances of output signal-to-interference-plus-noise-ratio (SINR) and bit error rate (BER) are improved.The least square constant modulus algorithm (LSCMA) is easy to capture the interference signals in the system that has several constant modulus signals. In order to overcome this shortage, a linearly constrained LSCMA multiuser detection algorithm is proposed. The spreading code of the desired user is used to impose the linear constraint on the conventional LSCMA, which makes the algorithm converge to the desired user. The proposed algorithm can solve the problem of the interference capture, and the convergence is not affected by the step size. Simulation results demonstrate that the proposed algorithm has faster convergence rate and better ability against MAI than the conventional LSCMA.The projecting approximation subspace tracking with deflation (PASTd) algorithm has slow convergence rate. To overcome this problem, an improved PASTd subspace tracking algorithm is proposed, and it is applied to the subspace-based MMSE semi-blind multiuser detector for adaptive subspace estimation, which enhances the convergence rate of the algorithm. A linear conjugate MMSE semi-blind multiuser detector based on the subspace method is designed by introducing the idea of the linear conjugate to the subspace-based MMSE semi-blind multiuser detection. In order to solve the problems of computational complexity and high hardware cost because of introducing the high-dimension matrix, a new signal subspace is defined, and the orthogonal projection approximation subspace tracking (OPAST) algorithm is used to estimate the signal subspace adaptively. The proposed algorithm not only improves the output SINR and BER performance, but also achieves a substantial reduction of the computational complexity.The spatio-temporal version of the subspace-based MMSE semi-blind multiuser detection is proposed with the combination of the subspace-based MMSE semi-blind multiuser detection and antennas array. By the subspace tracking approach, the signal steering vectors can be resolved, which can be efficiently computed at a comparable cost with that of the classic method. The proposed approach exploits the spatial characteristic of users, increases the ability against MAI of users. Simulation results show that the proposed algorithm not only has faster convergence rate but also achieves better output SINR and BER performance.A subspace constrained MOE semi-blind multiuser detection is proposed for uplink where the base station receiver has the knowledge of the spreading sequences of all the users within the cell, but not that of the users from other cells. A semi-blind multiuser detector based on the MOE criterion is proposed by exploiting the spreading codes of the users within the cell. Then, two modified adaptive algorithms such as the subspace constrained RLS algorithm and the subspace constrained least mean square (LMS) algorithm are proposed to combat the performance deterioration induced by the noise in the practical system, which is used to obtain the MOE weight vector adaptively. To reduce the computational complexity, the improved PASTd algorithm is developed for adaptive signal subspace estimation. This proposed detector suppresses the interferers, mitigates the noise, offers fast convergence rate, and improves the output SINR performance consistently.A seimi-blind MOE multiuser detection algorithm based on Generalized Sidelobe Canceller (GSC) structure is proposed. A linear conjugate MOE semi-blind multiuser detector is designed by introducing the idea of the linear conjugate to the MOE semi-blind multiuser detector. According to the GSC structure, a subspace-based weight vector is defined and the subspace method was combined with the adaptive algorithms to obtain'the weight vector adaptively. The proposed algorithm suppresses the interferers, mitigates the noise, and exploits the information contained in the pseudoautocorrelation of the observables. Thus the output SINR and BER performances are improved. The improved PASTd algorithm is developed for adaptive signal subspace estimation to reduce the computational complexity.