Skip to Main Content
Space-time coding promises greater insensitivity to wireless channel fading and improved channel capacity. In the current work, decoders based on the minimum mean squared error (MMSE) criterion are investigated. Two different MMSE-based decoders are developed based on knowledge about the residual error statistics. The performance of the decoders are evaluated in terms of achievable diversity levels for spread (so called "wideband" signaling) and non-spread ("narrowband") signaling as well as for single user and multi-user systems. The effects of multiple receive antennae is explicitly considered. Based on the pairwise symbol error rate, MMSE decoder optimized codes are found via exhaustive search. By evaluation of the union bound, it is seen that the MMSE decoder optimized codes offer good performance and ML decoder optimized space-time block codes generally maintain their good performance even when decoded through a MMSE decoder.