Blind channel estimation for MC-CDMA systems with long spreading codes

This paper develops the blind channel estimation for multicarrier code division multiple access (MC-CDMA) systems with long spreading codes. The channel impulse response plays an important role in estimation of the transmitted symbol. However, due to the long spreading codes, the time-varying effect makes the channel estimation difficult. The problem can be solved by utilizing the channel information matrix of the desired user which is computed by the second-order statistics of the despreading signal vector. It is analyzed that the novel channel information matrix is unbiased and is more efficient than that proposed by Liu [IEEE International Conference on Acoustics, Speech, and Signal Processing, 2000, ICASSP 2000, vol. 5, pp. 2909–2912] in the case of finite samples. With the channel information matrix, the channel impulse response can be estimated by the least-square method. Computer simulations show that the proposed method not only has smaller mean-square error of channel estimation but also requires fewer computations than Liu's method.