Abstract: According to the time-varying characteristic of information transmission channel between sensors distributed on remote and wild regions,which transmit data by existing speech channel in Early Warning System, a new method is proposed for adaptively adjusting transmission parameters in sensor networks based on channel sensing and monitoring. The method solves the problems of the sensing data loss and delay caused by connecting lines frequently dropped in low SNR (Signal to Noise Ratio).Firstly, we adopt a blind SNR estimation algorithm to sensing the current channel quality during the packets transmission period and obtain the observed values by integrating the estimation results with weighted sliding-window.Then , we establish a discrete adjusting transmission parameter model based on Kalman Filtering principle, which can adaptively adjust the transmission rate and power according to the changes in the channel quality while maintaining the link.Finally, experiments results show, in condition of wild field emplaced sensor network’s time-varying channel, this method can guarantee the information transmission’s instantaneity and reliability.