Adaptive Channel Estimation for OFDM Systems Using CIR Length Estimate

Abstract

This letter addresses adaptive channel estimation for OFDM systems to reduce computational complexity. There are numerous methods suggested for low-complexity channel estimation, but some of them still suffer from a considerable complexity or performance degradation due to the channel energy leakage. We present an adaptive channel estimator based on the channel impulse response (CIR) length estimate to reduce the complexity without any performance degradation. We first introduce a simple and efficient method based on the sequential hypothesis test (SHT) to estimate the CIR length and then apply it to the adaptive implementation of the SVD-based LMMSE estimator. Simulation results demonstrate that the SHT method performs well with very low complexity and that the adaptive estimator not only reduces the complexity significantly but also provides better performance than the fixed estimators.