Chip-level beamforming vs. symbol-level beamforming in coherent CDMA adaptive antenna array systems

In the coherent CDMA systems with adaptive antenna arrays, two representing architectures for the spatial-temporal CDMA demodulator are demonstrated based on the method of weight vector multiplication, which are chip-level beamforming and symbol-level beamforming. Depending on the architecture, the implementational complexity differs by 1.5~2.5 times. Moreover, the performance due to the method of searcher beamforming and finite word-length (FWL) effects is tightly related to the architecture of the spatial-temporal demodulator. Various beamforming algorithms are also compared in their performance under each architecture