This paper proposes an adaptive candidate selection scheme in the QRM-MLD algorithm for MIMO detection. The QRM-MLD is a near-ML detection algorithm which can achieve a tradeoff between the BER performance and the computational complexity in the MIMO systems. In this paper, we adopt an adaptive candidate selection scheme into the QRM-MLD. First, similarly to the conventional QRM-MLD, the proposed detection applies a fixed number of the survived branches to achieve a near-ML performance. Next, in order to evaluate the reliability of the survived branches in each detection layer, we introduce a ratio function of the path metric to the smallest path metric among the survived branches. The survived branch with lower reliability has less children nodes as the candidates in the next detection layer, which can avoid a large amount of the path metric evaluations and sorting. Hence, the complexity of the proposed detection should be low. Numerical results exhibit that the proposed scheme achieves the near-ML performance with lower complexity compared to the conventional QRM-MLD.

This paper proposes an improved lattice-reduction aided (LRA) MMSE detection based on the Gram-Schmidt (GS) procedure. The proposed detection reduces the column vectors of the MIMO channel matrix using the LLL algorithm and the GS procedure to create mutually purely orthogonal column vectors of the reduced channel matrix. Then the decision boundary becomes the same as that for the ML detection. Compared to the conventional LRA MMSE detector, the proposed detector achieves much closer BER performances to those for the ML detector in both the 4 × 4 MIMO and the 8 × 8 MIMO systems.

Lattice reduction aided (LRA) detection with the LLL algorithm has been investigated to achieve the MIMO channel capacity with low computational complexity. This paper proposes an improved LRA-MMSE detection using the forward and the backward reduction LLL algorithms for the 8times8 MIMO to obtain more reliable estimates of the transmitted signals. The proposed method achieved much better BERs than the conventional LRA detection. The transmit signal energy could be saved about 3.5 dB at BER=10-5, and the BER slope became much steeper, which is almost the same as that with the maximum likelihood detection.

This paper proposes a combined forward and backward lattice reduction scheme to improve estimation of transmitted signals in lattice-reduction aided (LRA) MMSE MIMO systems. The cross-correlation between the forward and the backward transform matrices is also analyzed to make sure that the proposed detector achieves more reliable estimates. The proposed detector achieves better BER performance than the conventional LRA detector. The achieved BER curve is improved by 3.5 dB at BER=10-5over the curve of the conventional LRA MMSE detector in the 8times8 MIMO.

Recently, lattice reduction aided (ERA) detectors have been introduced into Vertical Bell-Labs Layered Space-Time (V-BEAST) systems to obtain nearly optimal bit error rate (BER) performance for only small additional complexity. In this paper, the layer error characteristics of LRA-V-BLAST detectors are analyzed and compared with those of conventional V-BLAST ones. Two important conclusions are drawn for the LRA-V-BLAST detectors. First, the variation of their mean square error (MSE) within each detection iteration is not as large as in conventional V-BLAST detectors. Second, thanks to lattice reduction there exists an inherent sub-optimal detection order from the last to the first layer. These conclusions allow LRA-V-BLAST detectors to avoid optimal ordering to further reduce the complexity. LRA-V-BLAST detectors without optimal ordering are shown to obtain almost the same BER performance of LRA-V-BLAST detector with optimal ordering

In this paper, we propose two minimum mean square error (MMSE)-based multiuser detectors, called the combined MMSE and maximum likelihood (MMSE-ML) multiuser detector and the combined MMSE and successive interference cancellation (MMSE-SIC) multiuser detector, for an uplink space-time block coded orthogonal frequency division multiplexing (STBC-OFDM) system. Both detectors provide significantly improved bit error rate (BER) over the conventional MMSE multiuser detector while requiring the same complexity order. Moreover, the combined MMSE-ML detector exhibits better BER than the combined MMSE-SIC, particularly, in a system with a small number of users.

This paper compares the following two maximum likelihood sequence estimation schemes incorporating carrier phase estimation for coded MPSK: (1) the respective-state channel estimation and (2) the multiple differential detection. This paper confirms by analysis and computer simulation that both schemes have similar performance in the absence of intersymbol interference.

This paper studies a modulation method for a wideband code division multiple access (W-CDMA) system in order to enhance its performance. The modulation method studied here introduce bi-orthogonal modulation instead of multicode transmission in order to suppress the envelope variation which comes from code division multiplexing in multicode transmission. Bi-orthogonal modulation studied here selects a single code in both in-phase and quadrature components, and leads to relaxation of the requirement of HPA linearity. This relaxation enhances the implementation of the mobile station (MS) with smaller size or lower power consumption. Computer simulation results show that the power level of spurious emission in the adjacent channel is reduced by approximately 8 dB with use of bi-orthogonal modulation, while keeping the required E/sub b//I/sub o/ almost the same as multicode transmission.

This paper presents a new scheme of direct conversion phase shift keying (PSK) receivers applicable to time division multiple access (TDMA) land mobile communication systems. In the scheme, a matched filter in each I and Q channel is divided into two parts in front and at the back of a baseband amplifier to reduce baseband noise power and improve sensitivity. A rational design method to divide the filter is proposed; the filter cell that has narrowest equivalent noise bandwidth in the filter is chosen as the back part, while the remainder constitutes the front part. Moreover, memories are employed to eliminate DC offset in baseband signals. Experimental results show that, when applied to Personal Handy Phone System (PHS), a receiver employing the scheme achieves sensitivity comparable to that of a super-heterodyne receiver in continuous and burst reception.

This paper proposes an adaptive maximum-likelihood sequence estimation (MLSE) by means of combined equalization and decoding, i.e., adaptive combined MLSE, which employs separate channel estimation for respective states in the Viterbi algorithm. First, an approximate metric including channel estimation is derived analytically for this proposed adaptive combined MLSE. Secondly, procedures to accomplish blind equalization are investigated for the proposed MLSE. Finally, its excellent BER performance on fast time-varying fading channels is confirmed by computer simulation, when the proposed MLSE operates as a blind equalizer.>

This paper proposes an adaptive equalization scheme in dual diversity reception and assesses its performance by computer simulation over fading multipath channels. This scheme is composed of adaptive decision feedback equalizers (DFEs) in each diversity branch, a least-squares combiner (LSC) and a decision device. In the LSC, the output signals of DFEs are multiplied by controllable weights based on mean square error criterion and summed. The output of the LSC is applied to the decision device to obtain a receiver estimate of the transmitted sequence. The receiver estimate is applied to DFEs as the input of the feedback section and the reference of the tap coefficient-update algorithm. In the DFEs and the LSC, the tap coefficients are updated independently according to the Kalman algorithm. Such an arrangement of the scheme reduces the computational complexity compared with a conventional DFE transversal-combining diversity reception scheme. Simulation results confirm the effectiveness of this proposed scheme over fading multipath channels with transmission bit rate of 40 kbps and the maximum Doppler frequency of 160 Hz observed in 1.5 GHz-band digital mobile radio communication systems.

This paper proposes an adaptive maximum-likelihood sequence estimator (MLSE), which is a suboptimum approximation to the adaptive maximum-likelihood detector and is capable of tracking fast time-varying intersymbol interference (ISI) channels. The adaptive MLSE consists of a channel estimator and an MLSE implemented by the Viterbi algorithm (VA). The novel feature of the proposed adaptive MLSE is a channel estimation scheme, where the channel estimation is accomplished for each state in the VA along the surviving path connected to each state. This MLSE makes it possible to estimate a channel impulse response without an influence of a decision delay inherent in the VA, and the performance and complexity of the proposed procedure is controlled by the memory length of the VA. The bit error rate performance of the proposed MLSE is confirmed by experimental results. It is shown that the proposed MLSE has a capability of excellent tracking performance in a severe environment caused by fast time-varying ISI, for example frequency selective multipath fading in digital mobile radio communications.>

The effects of phase jitter and cycle slipping of the recovered carrier upon unique word (UW) missed detection in QPSK systems are investigated. Schrempp et al. have already presented the relationship between the UW missed detection probability (UWMP) and the bit error probability, which is simple and useful as long as the phase reference is noiseless. However, when the authors recently made some theoretical and experimental assessments of UW missed detection, they found that the above disturbances of the recovered carrier significantly affected the UW missed detection performance. Specifically, the actual UWMP was often worse than that predicted hy Schrempp et al. In fact, a bottoming effect takes place in the UWMP versus UW length characteristic. In this paper, after obtaining the conditional symbol and bit error probabilities for the recovered carrier with some phase error, we discuss the nature of phase fluctuation and cycle slipping for the recovered carrier. Consequently, we derive general expressions for the UWMP in both CPSK and DCPSK transmissions. The UWMP's are numerically calculated, and are then compared to our experimental data.