Decision feedback multiuser detection: a systematic approach

We consider interference rejection for code-division multiple-access (CDMA) channels using decision-feedback detectors. Performance of zero-forcing and the minimum mean square error (MMSE) multiuser decision-feedback detectors is studied, and the full and partial feedback approaches are investigated. We also derive the maximum-likelihood detector based on the white noise CDMA channel model. Comparison among the decision-feedback detectors, the two-stage receivers, the conventional detector, and the linear decorrelating detector is undertaken for a four-user bandwidth-efficient asynchronous CDMA channel. We find that the decision-feedback detectors compare favorably to other suboptimal methods and maintain good performance under diverse channel conditions. This conclusion is further strengthened by performance analysis of the decision-feedback detector for a two-user synchronous CDMA flat Rayleigh fading channel. >

An analytical framework for noncoherent decision feedback detection is introduced while considering nonorthogonal binary modulation (NBM) over the synchronous Gaussian K-user channel. Following the key idea of noncoherent decorrelating decision feedback detection (NC-DDFD) proposed previously, a K-parameter class of NC-DDFDs is defined. The symmetric energy measure is defined as the worst-case (over users) asymptotic effective energy for noncoherent detection. Without making any simplifying assumption about error propagation, the NC-DDFD that optimizes symmetric energy among the K parameter class of detectors is derived. Since the NC-DDFD based on the generalized likelihood ratio test (GLRT) belongs to the K-parameter class of NC-DDFDs, the optimum NC-DDFD outperforms the GLRT based NC-DDFD in symmetric energy. Like the latter detector, the optimum NC-DDFD does not require the knowledge of the energies or phases of any of the users' transmissions. Exact expressions for symmetric energy and upper and lower bounds for symbol error rate (SER) and asymptotic effective energies are obtained for the optimum NC-DDFD. Rules for ordering users are obtained that guarantee that the optimum NC-DDFD can user-wise outperform a parallel bank of post-decorrelative GLRT detectors.

This paper considers the joint design of bit loading, precoding and receive filters for a multiple-input multiple-output (MIMO) digital communication system employing decision feedback (DF) detection at the receiver. Both the transmitter as well as the receiver are assumed to know the channel matrix perfectly. It is well known that, for linear MIMO transceivers, a diagonal transmission (i.e., orthogonalization of the matrix channel) is optimal for some criteria. Surprisingly, it was shown five years ago that for the family of Schur-convex functions an additional rotation of the symbols is necessary. However, if the bit loading is optimized jointly with the linear transceiver, then the rotation is unnecessary. Similarly, for DF MIMO transceivers, a rotation of the symbols is sometimes needed. The main result of this paper shows that for a DF MIMO transceiver where the bit loading is jointly optimized with the transceiver filters, the rotation of the symbols becomes unnecessary and, consequently, also the DF part of the receiver is not required.

Existing work on decision feedback detection (DFD) of differential space-time modulation (DTSM) were largely confined to diagonal constellations. In other works considering nondiagonal constellations, the temporal variation of fading in the DTSM supersymbol duration was ignored. In this paper, we take a close look at the DFD receiver structure for nondiagonal DSTM in temporally correlated fading. We identify the constellations of which the error performance is not influenced by ignorance of fading variation. In addition, by analyzing the error performance, we demonstrate that existing DFD, is however, fundamentally mismatched to other constellations in fast fading. A major conclusion is that diagonal design still plays an important role for the efficacy of DFD for nondiagonal constellation in fast fading.

A low-complexity decision feedback (DF) detection algorithm with constellation constraints (DFCC) is proposed for multiple-antenna systems. An enhanced detection and interference cancellation is achieved by introducing the constellation constraints (CC). For time-varying channels, the proposed receiver updates the filter weights using recursive least squares (RLS) algorithm. This highly efficient detector is also incorporated in a multiple branch (MB) structure to achieve a higher detection diversity order. Simulations show that the proposed DFCC technique has a complexity as low as the conventional adaptive DF detector while it achieves a significant performance improvement. (5 pages)

RAKE receivers suffer degradation in performance under fast fading due to errors in channel state estimation. The wide-sense stationary uncorrelated scatterer channel model lends itself to a time-frequency canonical representation, and provides additional diversity in the frequency axis. Using the joint multipath-Doppler diversity model based on this canonical representation, we propose a nonlinear time-frequency decorrelating decision-feedback multiuser detector for synchronous CDMA which utilizes decisions of the stronger users when forming the decisions for the weaker ones. Analytical and simulation results based on realistic fast-fading assumptions demonstrate that the proposed decision-feedback multiuser detector promises substantially improved bit-error-rate performance compared to the time-frequency decorrelating multiuser detector for the weaker users, while the error probability of the weakest user approaches the single-user bound as the interferers grow stronger. The proposed detector exhibits strong resistance to the near-far effect for all users and the time complexity per bit is linear in the number of users.

We investigate detection structures for a DS-CDMA multiuser uplink using codes with a period of several symbols, named long codes. We first propose a multirate description of the uplink. Then based on this formalism we revisit a number of popular receiver structures: RAKE receivers, minimum mean square error linear and decision feedback detectors.

A fast optimal algorithm based on the branch and bound method, coupled with an iterative lower bound update, is proposed for the joint detection of binary symbols of K users in a synchronous correlated waveform multiple-access (CWMA) channel with Gaussian noise. The proposed method significantly decreases the average computational cost and the decision feedback detector (DFD) is proved to be a first order approximation to the optimal algorithm. Furthermore, a suboptimal algorithm using the probabilistic data association (PDA) is also proposed. It is shown that the suboptimal algorithm gives a near-optimal performance at a computational cost of O(K/sup 3/).

A multiuser decision-feedback detector (DFD) for direct-sequence code-division multiple-access (DS-CDMA) is proposed for fast-fading channels. The proposed modified minimum mean-squared error (MMSE)-DFD (termed precombining MMSE-DFD) is shown to be independent of the instantaneous channel coefficients. Hence, it can be adaptively implemented unlike the standard MMSE-DFD (termed postcombining MMSE-DFD), where it is known to break down in fast-fading channels due to the severe tracking problems. A two-stage MMSE-DFD is also derived which consists of two cascaded MMSE-DFDs, each performing parallel or successive interference cancellation prior to multipath combining. The performance of the various MMSE-DFDs is investigated and compared to the linear MMSE detector (i.e. no feedback) over different fading-channel models, using the Gaussian approximation. The numerical results show that, in all cases, the postcombining receivers perform significantly better than the corresponding precombining receivers. Finally, a simple adaptive implementation for the two-stage detector is derived based on the normalised least-mean-square (NLMS) algorithm. The performance of this two-stage receiver is examined using both the Gaussian approximation and Monte Carlo simulations.

The large system limits of the probability of error for multiuser decision feedback detectors (DFDs) are evaluated for synchronous CDMA. Both successive (S-) and parallel (P-) DFDs are considered, where the filters are optimized according to the minimum mean squared error (MMSE) criterion. A comparison of the analytical results with simulation shows that the large system results accurately predict the performance of systems with a spreading gain /spl ges/128 over a wide range of error rates. The results also show that the P-DFD error rate approaches the single-user bound at high signal-to-noise-ratios (SNRs).

This paper describes a new approximate maximum-likelihood (ML) MIMO detection approach by studying a Lagrangian dual relaxation (LDR) of ML. Unlike many existing relaxed ML methods, the proposed LDR employs a discrete domain for the problem formulation. We find that the proposed LDR exhibits an intriguing relationship to the lattice decoders (LDs) and the lattice reduction aided (LRA) detectors, both of which have caught much attention recently. Specifically, regularization in LDs, which was proposed to mitigate out-of-bounds symbol effects, can alternatively be interpreted as a way to constrain the symbol decision within bounds in a Lagrangian sense. We handle the LDR problem by using a projected subgradient method. The resultant method may physically be viewed as an adaptive regularization control in which a sequence of LDs are involved. Based on this newly developed insight, we propose two additional iterative LDR-based detectors using LRA decision-feedback (DF) and "lazy" DF. By simulation results, we show that the LDR LRA-DF and lazy-DF detectors yield better symbol error rate performance than the MMSE-regularized LRA-DF and DF detectors, respectively, where the SNR gaps can be more than 3dB.

The large-system performance of iterative multiuser decision-feedback detectors (DFDs) is studied for synchronous coded direct-sequence code-division multiple access. Both successive and parallel demodulation of users are considered. The filters are optimized according to the minimum mean-squared error criteria, assuming perfect feedback. We first consider Viterbi decoding with hard decision feedback, and compute union bounds on the large-system error rate. We then consider maximum a posteriori (MAP) decoding with soft decision feedback, and evaluate the error rate semianalytically by assuming the log-likelihood ratios computed by the MAP decoder are Gaussian random variables. Performance is studied numerically as a function of noise level, spectral efficiency, and code rate. Results show that soft decision feedback gives substantial gains relative to hard decision feedback. At moderate spectral efficiencies (users divided by bandwidth expansion less than 0.9), the iterative DFDs with soft decision feedback based on a posteriori probabilities can achieve near-single-user performance at an E/sub b//N/sub 0/ close to the large-system capacity bound.

In this paper, we design an implementation of a kind of reduced-complexity maximum likelihood (ML) detector for spatial modulation (SM) deployed in a multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) system. Based on National Instruments Universal Software Radio Peripheral (NI USRP) platform, two USRP Reconfigurable Input / Output (USRP-RIO) devices are used as the transmitter and the receiver. Error performance comparisons between the reduced-complexity ML detector and conventional ML detector are made in an indoor environment, as well as the comparison between SM scheme and space-frequency block code (SFBC) scheme. 6 different pictures are exploited as transmitted data, and are reconstructed in the receiver side, so the performance of each scheme is taken on the reconstructed pictures. Both of the calculated bit error ratio (BER) and the reconstruction result of each transmitted pictures, validate the highly consistent error performance of HL_ML detector and conventional ML detector, as well as the error performance superiority of SM over SFBC.

In this paper, we investigate the influence of time-varying channels on the bit-error-rate (BER) performance for distributed space-time block coded system. We make use of pilot symbol assisted modulation (PSAM) to estimate the time-varying channel coefficients. We assess the performance of different reception methods, one of which is our newly considered receiver called cooperative maximum likelihood detector. The others are maximum likelihood detection in [12], Alamouti's receiver, zero-forcing detection and decision-feedback detection. Our results show that all detection methods, except cooperative maximum likelihood detection, achieve nearly the same BER performance over time-varying channels, unlike the results obtained in (Vielmon et al., 2001) by assuming perfect channel estimation. In many cases, cooperative maximum likelihood detection performs better by about 5dB due to the diversity gain. In addition, we have found that the time-varying nature contributes to the error flooring effect.

In this work, arbitrarily tight upper and lower bounds to the asymptotic multiuser efficiency (AME) are derived for K asynchronous users and decision-feedback multiuser detection. Because the multiple-access interference is noncausal, the precursor interference may be canceled out by making tentative decisions via conventional or decorrelated decisions, or by the zero-forcing feed-forward equalizer (ZF-DFE). We demonstrate that decision-feedback multiuser detection with conventional tentative decisions produces low near-far resistance and that with decorrelated tentative decisions uniformly improves performance at the expense of a long decoding delay. The ZF-DFE shows low AME compared to decision-feedback detection with decorrelated tentative decisions. In order to use all strong users' final decisions, we propose a successively decorrelated decision-feedback technique for an asynchronous CDMA system. This detector is shown to uniformly outperform the decorrelated detector and stays close in performance to the optimum multiuser receiver.