A novel channel fault-tolerant localization algorithm based on maximum-likelihood estimation is proposed for localization in a multi-hop wireless sensor networks with channel fading. The received measurement of source signal energy from local sensors is compressed and quantized into binary data and transmitted to the fusion center via multi-hop Rayleigh fading channels. The likelihood function of binary observation is derived by using wireless channel statistics and decoding scheme. The localization of target is achieved by maximum-likelihood estimation. Furthermore, the Cramér-Rao lower bounds for the estimates are derived. Simulation results show the effectiveness of the proposed algorithm.