SVD pre/post-RAKE with adaptive trellis-coded modulation for TDD DSSS applications

Since reciprocal channels can be found in time-division duplex (TDD) systems, TDD direct-sequence spread-spectrum systems are capable of implementing a pre-RAKE structure as a predistortion technique. In this paper, we propose a singular value decomposition (SVD) pre/post-RAKE in which the RAKE combiner is utilized at both the transmitter and the receiver. If the combining weights of the pre- and post-RAKE are determined by the singular value decomposition, then parallel subchannels can be established in the eigenspace. We obtain the optimal information capacity by applying the water-filling solution to the SVD pre/post-RAKE and investigate the capacity behavior at an asymptotic high SNR. The SVD pre/post-RAKE provides remarkable capacity improvement without increasing the average transmit power or frequency bandwidth, since multiple data streams can be reliably transmitted through the parallel subchannels. For practical implementation, the adaptive trellis-coded modulation technique with a bit and power-allocation algorithm is applied to the SVD pre/post-RAKE. Compared with other RAKE structures, the SVD pre/post-RAKE provides a significant performance gain in high data rate communication.