Low Complexity Block Processing Algorithms for Adaptive Channel Estimation in OFDM Systems

For pilot-aided channel estimation in orthogonal frequency-division multiplexing (OFDM) systems, adaptive algorithms enable tracking of a time-varying channel. Using several subcarriers for the update step in the adaptive algorithm, i.e., performing block processing, improves the robustness and convergence speed of the adaptation. To decrease the complexity of these algorithms, simplified versions are deduced as well as versions with real-valued filters. In the latter case the number of required arithmetic operations can be reduced significantly not only in the adaptive algorithm but also in the filtering process itself, which represents a dominating part regarding the complexity due to the huge amount of subcarriers of practical OFDM transmission schemes. Simulation results show that gains w.r.t. the convergence behavior can be achieved by real-valued filters compared to complex-valued filters - even in scenarios where real-valued filters are theoretically suboptimum.