Abstract: FIM(fisher information matrix) and CRB(cramer rao bound) for complex channel parameters estimation in Gaussian signal model and in deterministic signal model are analyzed separately in this paper. Furthermore, analysis of FIM and CRB for complex channel estimation is implemented when Gaussian noise power is assumed unknown. As for the singularity of FIM for blind channel estimation, CRBs under different constraints are analyzed. In the end, blind channel estimation algorithms are simulated and compared with CRB in deterministic signal model and in Gaussi an signal model separately.

Key words: blind channel estimation, root mean squared error for channel estimation, cramer rao bound