Title :
Linear MIMO equalization for CDMA downlink signals with code reuse
Author :
Mailaender, Laurence
Author_Institution :
Lucent Technol. Bell Labs., Holmdel, NJ, USA
Abstract :
We consider a code-division multiple-access downlink using multiple-input multiple-output (MIMO) transmission/reception and spatial reuse of orthogonal codes to achieve high throughput. We focus on the equalizer design for such systems when the channel has significant temporal dispersion leading to orthogonality loss. A chip-level minimum mean square error (MMSE) linear equalizer is used to simultaneously correct the temporal dispersion while providing separation of the spatial signals. Closed-form linear MMSE solutions for MIMO systems are given in both the time and frequency domains, and results are compared. An expression for the signal-to-interference-plus-noise ratio after equalization and despreading is found and used as the basis for the evaluations. We investigate the required equalizer lengths and impact of oversampling, and we compare two forms of adaptive least-mean-square implementation on 4 × 4 time-invariant MIMO channels.
Keywords :
MIMO systems; channel coding; code division multiple access; dispersive channels; equalisers; least mean squares methods; radio links; time-frequency analysis; wireless channels; CDMA downlink signal; code division multiple access; code reuse; dispersive channel; equalizer design; frequency-selective channel; linear MIMO equalization; minimum mean square error linear equalizer; multiple-input multiple-output reception; normalized least mean square method; orthogonal code; oversampling; signal-to-interference-plus-noise ratio; time-frequency domain; 3G mobile communication; Downlink; Equalizers; Fading; Interference; MIMO; Multiaccess communication; Receivers; Receiving antennas; Throughput; Frequency-selective channels; multiple-input multiple-output (MIMO); normalized least-mean-square (NLMS); oversampling;
Journal_Title :
Wireless Communications, IEEE Transactions on
DOI :
10.1109/TWC.2005.853887