Low-Complexity Dynamic Pilot Allocation for Channel Estimation in MIMO-OFDM Systems with Spatial Multiplexing

Inhalt:
Dynamic pilot allocation for Discrete Fourier Transform (DFT)-based channel estimation in MIMO-OFDM system with spatial multiplexing has been shown to significantly improved symbol error rate (SER) performance compared to fixed pilot allocation. However the exhaustive search for optimum allocation leads to very high complexity. In this paper, we investigate a MIMO iterative pilot search (MIPS) algorithm applied with different MIMO-OFDM receivers (ZF-linear, ZF-SIC, MMSE-linear, MMSE-SIC), which significantly reduces the complexity of the search for optimum pilot patterns, and makes the dynamic algorithm more feasible for practically useful number of subcarriers and antennas. Furthermore, the mathematical derivations of the optimum pilot search are discussed briefly including the derivation of the signal to interference plus noise ratio (SINR) in the presence of channel estimation errors and approximations of error probabilities. The effectiveness of the new techniques is demonstrated by simulation of a MIMO-OFDM system. The SER performances of the algorithm based on MIPS are almost the same as the exhaustive search.