Channel estimation for orthogonal frequency division multiplexing (OFDM) systems operating over doubly selective channels is complicated by the existence of intercarrier interference (ICI).This paper extends the idea of estimating time-variant channel impulse response (CIR) via time-averaged CIR proposed by Hijazi,and proposes a low-complexity channel estimation scheme with ICI suppression based on Karhunen-Loeve basis expansion model (KL-BEM),which is in a more general form.In the proposed scheme,the KL-BEM coefficients are obtained through their numerical relationship with the time-averaged CIR.The estimation performance is improved by adopting an expectation-maximization (EM)-based iterative structure combined with a low-complexity ICI equalizer and canceller.Theoretical analysis and simulation results show that the proposed scheme achieves superior performance over the original one with lower computational complexity.
