Precise Point Positioning (PPP) has been considered a powerful method for GNSS data processing. The essential input products, such as precise satellite orbits and clocks, are provided within the International GNSS Service (IGS) with a sufficient quality for estimating receiver coordinates with centimeter level accuracy. However, the IGS satellite clocks enable users to estimate ambiguities only as float values. An additional product for satellite phase biases is necessary for an integer ambiguity resolution (PPP AR). Another approach is the backward smoothing algorithm utilizing already precise and converged parameters for improving those parameters estimated at previous epochs. All the three approaches for ambiguity estimation are compared and assessed in terms of advantages and disadvantages, achieved coordinates precision, and flexibility. The comparison are performed through a processing of GNSS data from selected IGS permanent stations during 30 days in 2018, and a processing of high rate GNSS observations of the station STRF in Greece collected during the seismic event occurred on October 25, 2018. The backward smoothing improved the float solution similarly like the PPP AR, and therefore can be considered an alternative approach providing easier implementation and no dependency on additional satellites products. We utilized two different products for phase biases in the PPP AR, namely Integer Recovery Clocks (IRC) provided by the Centre National d’Études Spatiales/Collecte Localisation Satellites (CNES/CLS) analyses center and Fractional Cycle Biases (FCB) which were estimated at the Geodetic Observatory Pecny (GOP) analyses. The IRC is based on the assimilation phase biases into satellite clocks, while the FCB products are distributed in terms of wide-lane and narrow-lane biases. A similar accuracy obtained from our comparison indicates an interoperability of products when using different strategies and even different software.