Abstract: Multipath is one of main errors in GNSS. Among the existing research, the maximum-likelihood (ML) methods are regarded as the best to mitigate multipath but computationally complex, especially when there are many multipaths and the number of multipaths is unknown. A multipath mitigation technique based on FIR model is put forward. The reference correlation function and the minimum mean square error principle is used to estimate the parameters and eliminate the multipath of received signals. The method is provided with simple implementation and high realtime processing capability, and its computation complexity is independent of the amount of multipaths. Simulation results show that it is superior to the MEDLL (Multipath Estimating Delay Lock Loop) technique based on ML.