Research On MIMO Detection And Transceiver Design

Multi-Input Multi-Output (MIMO) technique has attracted many interests due to its high spectra efficiency and transmission capacity in broadband wireless communications. However, the design of low complexity signal processing schemes capable of supporting data rates close to the MIMO capacity remains a major challenge. In this dissertation we do research on MIMO detection and transceiver design. The main contents are as follows,Propose an ordered group interference cancellation for quasi-orthogonal space time block codes by studying the post detection signal-to-noise ratio (SNR) of transmit symbols at each detection step of ordered interference cancellation; Propose a low complexity detection scheme, based on sequential Gaussian Approximation (SGA), called multi-level mapping SGA detection, for MIMO wireless communication systems with high-order QAM modulation. Due to the property of high-order QAM formats, the new scheme calculates the sequence weight in step of two bits, resulting in a complexity reduction from exponential complexity in the modulation order to linear complexity, compared to basic SGA detection; Propose a bidirectional detection method to prevent error propagation based on Gaussian approximation algorithm for MIMO systems. In the new method, transmitted symbols are estimated in parallel both in sequential order and inverse order, and the number of searching paths is determined by comparing the results of bidirectional detection, which can prevent error propagation with reduced complexity. The novel algorithm can be executed at low delay because of parallel computing, and shows good performance at low complexity; Propose a kind of MIMO-OFDM Turbo transceiver with good performance to suppress the co-antenna interference (CAI) in wireless communication. The proposed scheme reaches good tradeoff among the data rate, transmit diversity, and system performance by using Turbo iterative processing based on pre-designed pattern mapping for multi-stream MIMO OFDM. At the transmitter, pre-designed symbol-mapping pattern is used for each transmit antenna. At the receiver, a TWO-STAGE detector receiver, including MMSE in stage one and MAP in stage two, performs two successive decisions, achieved by a maximum a post-probability (MAP) detector and a soft-input soft-output (SISO) channel decoder through an iterative process. At each iteration, extrinsic information is exchanged between the detector and the decoder.Propose a reliability assessment (RA) for MIMO detection at the receiver, which can assess whether the output of MIMO detection is reliable or not. First, an analytical formula of the reliability assessment is derived to assess the reliability of the receiver's output, a simplified reliability assessment is also proposed and a complexity-reduced method is further given by adjusting the threshold of reliability assessment, with some potential performance loss compared to the original method. Next we develop a combined detector with simple receiver, advanced receiver and reliability assessment module as follows: 1) the initial estimate of transmit symbols is obtained by the simple receiver, assessed by RA in the reliability assessment module; 2) if the estimate satisfies the RA, we consider the estimate as the final result; otherwise, the estimate will be through the advanced receiver for further improvement.Propose a more accurate approximation for L-value soft output of the systematic bits by using Gaussian mixture distribution (GMD) instead of Gaussian distribution, to obtain more reliable bit error rate (BER) predictions of Turbo codes based on extrinsic information transfer chart. Parameters pertaining to GMD are estimated via the expectation maximization (EM) algorithm. Numerical illustrations demonstrate that the proposed method can obtain reliable BER predictions of Turbo codes down to 10^-5.Propose an adaptive MIMO system based on the unified belief propagation (BP) detection. A BP detector composed of a serial of component detectors is first presented. With a unified design of the BP detector, a multi-rate MIMO system adaptive to channel variations is then developed, which aims at achieving the maximal throughput performance. The usage of the unified BP detector eliminates the multiple detector modules at the receiver in conventional multi-mode adaptive MIMO systems. Further, the linear complexity to space time block length and parallel operation of the unified BP detection provide the advantages of low complexity, low detection delay and smooth switching between different data rates. With the average post-detection SNR based switching mechanism, the proposed adaptive MIMO framework adaptively chooses the appropriate rate for the transmission in presence of spatial correlation.