Aiming at the problem of low reliability of real-time ambiguity fixation in kinematic-to-kinematic relative positioning based on single-frequency carrier phase. A smoothed INS-aided ambiguity resolution strategy for kinematic-to-kinematic relative positioning is proposed in this paper, to improve the universality of high precision positioning for low cost receivers. Firstly, a kinematic-to-kinematic relative positioning model was established based on double-difference (DD) pseudorange and carrier phase observations. Secondly, a Kalman filtering was applied to the inertial navigation data, and the DD observation equations were reconstructed using the prediction pseudorange by filtered INS-derived position. Meanwhile, the INS-derived position constraint equation was complemented. The high precision ambiguity float solutions and their covariance matrix were obtained, then LAMBDA algorithm was used to fix the ambiguity and Ratio test was conducted. Finally, the INS errors were feedback corrected by DGNSS/INS tightly coupled Kalman filtering results to maintain INS high precision navigation. The experimental results of testing data show that the method adopted in this paper can effectively increase the accuracy of ambiguity float solution, reduce the search space for ambiguity, and improve the success rate of ambiguity fixation. By comparing the results of without INS aided, with INS aided and with smoothed INS-aided, the success rate of the three methods were 92.3%, 99.1% and 99.6%, respectively, and finally achieve dynamic relative positioning in centimeters.