SPC03-2: A Robust Sampling Iteration Detection for Fast Flat Fading MIMO Channels

In this paper, a sampling based channel estimation as well as an iterative particle filter (PF) signal detection scheme for fast flat fading multiple-input multiple-output (MIMO) channels is proposed, which aims at achieving the required bit error rate (BER) performance under the scenario that the accurate channel estimation is difficult to attain. The channel estimation is comprised of two parts: the adaptive iterative update on the channel distribution mean and a regular update on the "adaptability" via pilot. In the detection procedure, the particle filter (PF) is employed to produce the optimal decision given the known received signal and the sequence of the channel samples, where an asymptotic optimal importance density is constructed, and in terms of the asymptotic update order, the parallel importance update (PIU) and the serial importance update (SIU) scheme are implemented respectively.