This paper proposes non-coherent multiple-input multi-ple-output (MIMO) communication systems employing per transmit antenna differential mapping (PADM), which generates an independent differentially encoded sequence for each of the multiple transmit antennas by means of space-time coding and mapping. At a receiver, the proposed PADM employs adaptive maximum-likelihood detection (MLD). The features of PADM are as follows: 1) it has excellent tracking performance for fast time-varying fading channels, because it can detect transmitted data without needing channel state information (CSI); 2) it can be applied not only to transmit diversity (TD) but also to spatial multiplexing (SM). In this paper, we analyze the adaptive MLD based on pseudo matrix inversion and derive its metric for data detection. In order to satisfy requirements on multiple transmitted sequences for the adaptive MLD, this paper proposes a mapping rule for PADM. Next, this paper describes a receiver structure based on per-survivor processing (PSP), which can drastically reduce the complexity of adaptive MLD. Finally, computer simulations confirm that the proposed non-coherent MIMO communication systems employing PADM have excellent tracking capability for TD and SM on fast time-varying fading channels.

This paper proposes a co-channel interference cancellation method for multiple-input multiple-output (MIMO) wireless communication systems. Maximum-likelihood multi-user detection (ML-MUD), which is one of the co-channel interference cancellation methods at a receiver side, has excellent bit error rate (BER) performance. However, computational complexity of the ML-MUD is prohibitive, because the ML-MUD must search for the most probable symbol vector from all candidates of the transmitted signals. We apply sphere decoding (SD) to the ML-MUD in order to reduce the computational complexity of the ML-MUD, and moreover we propose a modified version of the SD suitable for the ML-MUD. The proposed method extracts desired signal components from a received signal vector and a channel matrix decomposed the upper triangular form, and then performs the SD to the low dimensional model in order to detect the transmitted signals of the desired user. Computer simulation confirms that the proposed method can suppress the undesired signals and detect the desired signals, offering significant reduction of the computational complexity of the conventional method.

This paper proposes a novel channel estimation method for iterative equalization in MIMO systems. The proposed method incorporates co-channel interference (CCI) cancellation in the channel estimator and the channel estimation is successively performed with respect to each stream. Accuracy of channel estimation holds the key to be successfully converged the iterative equalization and decoding process. Although the channel estimates can be re-estimated by means of LS (Least Square) channel estimation using tentative decisions obtained in the iterative process, its performance is severely limited in a MIMO system because of erroneous decisions and ill-conditioned channel estimation matrix. The proposed method can suppress the above effects by means of CCI cancellation and successive channel estimation. Computer simulation confirms that the proposed channel estimation method can accurately estimate the channel, and the receiver with iterative equalization and the proposed method achieves excellent decoding performance in a MIMO-SM system.

This paper proposes a frequency-domain equalizer (FEQ) which utilizes not only guard interval but also redundancy in the frequency domain to eliminate inter-symbol and inter-carrier interferences. The proposed FEQ employs the hybrid criterion, i.e., the zero-forcing (ZF) criterion for compensating desired subcarriers and the minimum mean square error (MMSE) criterion for suppressing interference. The proposed Hybrid-FEQ achieves a good equalization performance, because it can suppress the noise enhancement caused by the ZF criterion with relatively small computational complexity exploiting soft-decision forward error correction (FEC). In this paper, we show its equalization performance and complexity compared with the conventional FEQs.

This paper proposes a likelihood estimation method for reduced-complexity maximum-likelihood (ML) detectors in a multiple-input multiple-output (MIMO) spatial-multiplexing (SM) system. Reduced-complexity ML detectors, e.g., Sphere Decoder (SD) and QR decomposition (QRD)-M algorithm, are very promising as MIMO detectors because they can estimate the ML or a quasi-ML symbol with very low computational complexity. However, they may lose likelihood information about signal vectors having the opposite bit to the hard decision and bit error rate performance of the reduced-complexity ML detectors are inferior to that of the ML detector when soft-decision decoding is employed. This paper proposes a simple estimation method of the lost likelihood information suitable for the reduced-complexity ML detectors. The proposed likelihood estimation method is applicable to any reduced-complexity ML detectors and produces accurate soft-decision bits. Computer simulation confirms that the proposed method provides excellent decoding performance, keeping the advantage of low computational cost of the reduced-complexity ML detectors.