Owing to the state transition and environment relativity of the communication link of wireless sensor networks（WSNs），the estimation and prediction of link-quality has become an important part of realizing reliable transmission of WSNs in smart distribution network communication.This paper presents the development of a link-quality estimation and prediction model of WSNs in smart distribution network communication.Then the main influencing factors of the model including disturbances of the background noise and signal strength attenuation are analyzed.The alpha-stable distribution is adopted to fit the background noise while developing a mathematical model on the disturbance of background noise.Meanwhile，the least square method（LSM）is used to estimate the environment parameters influencing the signal strength while developing a mathematical model for the disturbance of signal strength attenuation.Finally，it is shown by experiment that the proposed link-quality estimation and prediction model is able to accurately estimate and predict the actual link-quality.