Improved Semi-Blind Sparse Channel Estimation Based On Constant Modulus Constraint

Liping Dong,Qun Wan
DOI: https://doi.org/10.1109/WICOM.2010.5600256
2010-01-01
Abstract:Semi-blind channel estimation, as a combination method of blind channel estimation and training-based channel estimation, has an attractive improve on the performance of frequency spectrum efficiency. Enforcing the sparsity of channel impulse response as well as the constant modulus property of symbols, taking advantage of known training-based data, as well as unknown constant modulus symbols, an improved semi-blind channel estimation algorithm is proposed. Simulation results show that the proposed algorithm has better performance than the only training-based least-squares method and sparse component analysis method, and the required length of training symbols is much shorter, which means using less computational complexity to improve frequency spectrum efficiency.
What problem does this paper attempt to address?