A Low-Complexity Channel Estimation for OFDM Systems Based on CIR Length Adaptation

This paper addresses an adaptive implementation of frequency-domain channel estimators for orthogonal frequency-division multiplexing (OFDM) systems to reduce computational complexity. The linear minimum mean-square-error (LMMSE) estimator provides optimal performance at the cost of high computational load. There are numerous approaches to achieve low-complexity channel estimation, but many of them suffer from irreducible error floor due to channel energy leakage. In this paper we present an adaptive channel estimation based on channel impulse response (CIR) length estimation. We introduce a simple and efficient method for estimating the CIR length and then apply the method to the singular value decomposition (SVD)-based low-rank approximation. Simulation results demonstrate that the proposed method for CIR length estimation is highly efficient and that the adaptive implementation of the SVD-based estimator can reduce the complexity significantly when the channel delay spread is small and also achieves a considerable performance improvement in a time-varying channel.