Date Added: Jul 2011
This paper presents a subspace tracking based channel estimation utilizing Bi-iterative Least-Square (Bi-LS) rank adaptation method in sparse channel for OFDM systems. After modeling parameterized fading channel with sparse multipath, a general description of parametric channel estimation process based on subspace tracking is given. Then a simple Bi-iterative Least-Square (Bi-LS) method is discussed with its capability of rank adaptation and low computational complexity. Finally, numerical simulation results show that Bi-LS method has better performance not only in subspace tracking, but also in subspace rank updating compared with other channel estimation techniques, especially when the multipath number is time variant or the sampling timing synchronization jitters.