The extraction of high-precision ionospheric observables is essential for developing accurate ionospheric models. Current techniques for extracting ionospheric total electron content (TEC) include code ionospheric observables (CIOs), carrier phase ionospheric observables (CPIOs), carrier-to-code leveling ionospheric observables (LIOs), and precise point positioning (PPP) ionospheric observables (PIOs). However, these methods are impacted by multipath and noise interference. To address these limitations, we proposed a new approach: the PPP ambiguity resolution (PPP-AR) method, which utilizes code and carrier phase observations based on integer ambiguity for ionospheric observables extraction. In this study, we introduced the PPP models and PPP-AR process, and then we derived the mathematical formulas for the ionospheric observables of PPP-AR methods. To evaluate the effectiveness of our methods, PPP-float and PPP-AR experiments with GPS/Galileo/BDS observation from 12 stations over one week were conducted. Results showed that the performance of PPP-float was improved from 1.59, 1.25, and 2.24 cm to 1.21, 1.08, and 1.95 in the east, north, and up directions, respectively. The accuracy of ionospheric observables derived from PPP-float and PPP-AR was assessed by zero- and short-baseline experiments, showing that STD of single-differenced slant TEC (STEC) extracted by PPP-float were 1.4, 2.1, and 2.0 TECu for GPS, GAL, and BDS, respectively; while those were 0.05, 0.05 and 0.07 TECu for PPP-AR. Compared with float PIOs, the accuracy of fixed PIOs was improved by 96%.