For the existing link quality estimation with no consideration of the incomplete packet impact on link quality in wireless sensor networks

the packet loss was divided into two categories-perceived packet loss and non-perceived packet loss.A link quality estimation model under non-perceived packet loss has been proposed on the basis of the analy-sis of the causes for the non-perceived packet loss and the measurement of the non-perceived packet loss.When the model is being built

Kalman filter is introduced to denoise the acquired CCI（chip correlation indicator）.The cubic model and the least squares method are employed in the fitting relationship between CCI and non-perceived packet loss ratio.The experiment results show that the link quality estimation model is reasonable

and non-perceived packet loss ra-tio coming from this model is very close to the measured value.