An improved ZF-ML decoder for space time block codes over time selective channels

In this paper we propose a novel zero-forcing maximum likelihood (ZF-ML) decoder for symbol detection with reduced search space in Alamouti's space-time block coding (STBC) scheme over time-selective fading channels. Once the channel is time-varying in nature, the channel coefficients vary from one signaling interval to another and causes inter-transmit antenna interference(ITAI); the maximum likelihood (ML) decoder carries out exhaustive search and thus becomes computationally expensive. We evaluate the performance of proposed ZF-ML decoder as compared to the generalized simplified ML (GSML) decoder proposed by T. A. Tran and A. B. Sesay, and sphere decoder (SD) in time-varying channels assuming perfect channel state information(CSI) along with imperfect CSI, being more realistic, at the receiver. Simulation results suggest that the use of improved ZF-ML decoder in time-selective channel gives close to optimal results at much reduced complexity and improved results as compared to the GSML decoder having same computational complexity.