— In recent years, several approaches have been proposed aiming the optimal joint design of finite impulse response (FIR) multiple-input multiple-output (MIMO) transmitter and receiver. Assuming that the channel input is power constrained and that the channel model is a known FIR MIMO system, it is possible to design high performance transceivers in the presence of additive noise. Unfortunately, most optimal transceivers available so far, do not have fast implementation and simple design approaches. This work proposes a minimum mean square error (MMSE) design approach where constraints are incorporated into the structure of the FIR MIMO transmitter, in such a way that the resulting optimal MIMO transmitter, in the MMSE sense, has inherent fast algorithm implementation. In comparison with the general jointly optimized transceivers recently proposed, the new approach allows substantial reduction in the computational complexity at expense of mild performance degradation.
Guilherme Pinto, Paulo S. R. Diniz, Are Hjø