Dynamic pilot allocation with channel estimation in closed-loop multi-input-multi-output orthogonal frequency division multiplexing systems

Dynamic pilot allocation (DPA) for discrete Fourier transform (DFT)-based channel estimation in multi-input-multi-output orthogonal frequency division multiplexing (MIMO-OFDM) systems with spatial multiplexing can significantly improve the bit error rate performance compared with systems with uniform pilot allocation. However, the exhaustive search for optimum pilot allocation leads to very high complexity. The authors devise a multi-input-multi-output iterative pilot search (MIPS) algorithm applied with different MIMO-OFDM receivers (linear, successive interference cancellation (SIC) and maximum likelihood (ML)), which significantly reduces the complexity of DPA. Exact derivations are also given based on the receivers. They also propose a novel stacked vector quantisation technique to reduce feedback burdens for DPA in MIMO-OFDM system. Simulation results illustrate that the proposed MIPS algorithm with limited feedback can improve the performance of MIMO-OFDM systems.