Adaptive Space-Time Processing for Wireless CDMA

Abstract

We consider adaptive space-time processing for wireless receivers in CDMA networks. Currently, the 2D RAKE is the most widely used space-time array-processor which combines multipath signals sequentially, first in space, then in time. We introduce incremental processing improvements to arrive ultimately at a more efficient one-dimensional joint space-time (1D-ST) adaptive processor named STAR, the spatio-temporal array-receiver. STAR improves the receiver’s performance by approaching blind coherent space-time maximum ratio combining (ST-MRC). With blind quasi-coherent joint ST-MRC, STAR outperforms the 2D RAKE by improving the combing operation with 2 dB gain in SNR. With quasi-blind coherent joint ST-MRC, STAR can enhance channel identification while reducing significantly the pilot-power or -overhead fraction leading to a 1 dB gain in SNR. These gains translate to significant performance advantage for all versions of 1D-ST STAR over current 2D RAKE-type receivers.