Three-stage iterative detection of a MIMO-aided precoded AMR-WB speech transceiver

Abstract

A jointly optimised iterative source and channel decoding (ISCD) scheme invoking the Adaptive Multi RateWideband (AMR-WB) speech codec is proposed. More explicitly, the transceiver investigated consists of serially concatenated Recursive Convolutional (RSC) codes, a Unity Rate Code (URC) referred to as a precoder and the AMRWB speech codec, where the resultant bitstream is transmitted using Differential Space-Time Spreading (DSTS) and Sphere Packing (SP) modulation over narrowband temporally correlated Rayleigh fading channels. The convergence behaviour of the advocated scheme is investigated with the aid of Extrinsic Information Transfer (EXIT) charts, which suggests that potential performance improvements can be achieved by combining a serially concatenated precoder with the AMR-WB speech codec. The proposed system exhibits an E b/N0 gain of about 1.5 dB in comparison to the benchmark scheme carrying out iterative source- and channel-decoding as well as DSTS aided SP-demodulation, but dispensing with the precoder, when using I system = 4 system iterations. © 2009 IEEE.