A low-complexity soft-output Max-log-MAP (maximum a posteriori probability) detector for N2-QAM is exemplified for a 64-QAM 2timesM multiple-input-multiple-output (MIMO) system, exhibiting the following novel features: (1) hierarchical formulation of the exhaustive metric calculation to reduce the number of candidate tests by a factor of N per received symbol, and (2) multiplier-free implementation of the exhaustive search with 8-fold parallelization. The computational complexity is reduced by a factor of about 250, resulting in a chip area of 0.031 mm2using 65 nm CMOS.

For WCDMA base stations using adaptive antennas, we present a new space-time RAKE structure for uplink reception. Our approach combines short-term and long-term spatial and temporal channel properties using an eigenanalysis. By choosing dominant eigenbeams in time and space, the algorithm enhances interference suppression as well as spatial and temporal receive diversity. In contrast to previously introduced well-known receiver structures, the space-time eigenRAKE inherently adapts to different propagation environments and achieves higher spectral efficiency than other receivers, e.g., the beamformer RAKE. This is illustrated by Monte-Carlo simulations. Finally, we extend the proposed concept to the downlink, improving closed-loop downlink diversity compared to other proposals in standardization. Even though the feedback rate remains unchanged, additional antenna elements can be included to increase the antenna and diversity gains.

We study the utilization of adaptive space-time (ST) signal processing for mobile receivers in frequency-selective cellular radio channels with cochannel interference and correlated fading. First, we introduce a bad urban spatio-temporal channel model for the perspective of a mobile multiple antenna receiver, present mobile array structures and inter-element correlations, and investigate cochannel interference conditions. We then conduct a comparative study for two promising combiner-equalizer structures, viz. a temporal decision-feedback equalizer (DFE) with a preselection diversity switch and a ST-DFE, all using recursive least squares (RLS) adaptation. Our results indicate that both techniques yield performance gains with respect to a single antenna receiver, with the more expensive joint space-time method often strongly superior to other setups. Finally, we present an efficient and stable structure for implementing the ST-DFE, using a QR-decomposition type of RLS adaptation.

Parallel architectures for decision-directed adaptive recursive least-squares (RLS) equalizers based on a QR-decomposition (QRD) using square-root free Givens-rotations are derived that allow very high throughput for broadband applications. We apply an algorithmic optimization that leads to the use of additional fast adder trees together with a re-timing of the required arithmetic components to maximize throughput. A hybrid RLS/least-mean-squares equalizer is further proposed allowing even higher throughput. The complexity of the resulting equalizer structures is compared for an eight-tap decision-feedback equalizer in terms of an equivalent NAND gate count.

We investigate receive diversity for OFDM-based broadband communication systems. We consider the use of multiple antennas to combat time-selective fading and OFDM transmission in combination with an appropriate guard interval to suppress intersymbol interference. The receive diversity scheme is located in the frequency-domain and we derive different linear estimators, hereafter referred to as diversity equalizers. We show that the carrier-based solution for the optimum linear MMSE estimator-which is the least complex solution that maximizes the overall SNR is the well known maximum ratio combiner (MRC). Furthermore, we present a performance analysis for the proposed diversity equalizers and compare theoretical results to a simulated system which uses channel state information for MRC.

An extended quadratic residue (QR)-decomposition (QRD)-based RLS-algorithm is introduced, applicable for equalization, allowing an estimate of the transmitted symbol, when it is a-priori unknown. To achieve low-cost implementations strength reduction and square-root free Givens-rotations are applied. Several QRD-based RLS-algorithms are compared in terms of the number of mathematical operations. The QR-RLS-algorithm indicates lowest complexity and is always a good candidate, when a huge training overhead has to be avoided, resulting in an equalizer of less than 15 taps. Finally, the hardware complexity of a 13 MBaud-DFE for wireless ATM is estimated.