Convergence analysis of an adaptive channel estimation algorithm using cyclic prefix in multicarrier modulation systems

Multicarrier Modulation (MCM) gains growing interests in high data rate communications for both wire and wireless environment. The channel estimation is a crucial aspect in MCM systems. A novel channel estimation algorithm [6] exploiting the channel information contained in the cyclic prefix outperforms the existing adaptive equalization scheme [Z]. In this paper theoretical analysis is carried on the convergence of the adaptive channel estimation algorithm considering both channel noise and decision error. We prove that the algorithm is guaranteed to converge with proper loading. We also analyze several factors that affect the convergence rate such as data structure and loading function. Computer simulation shows that our analytical results are quite close to the simulation.