Synthetic Aperture Radar (SAR) sensors on-board satellites are very well suited for observing sea surface geophysical parameters such as ocean swell. But on a very large scale, SAR data are too sparse for deriving some global information. From the original work of Collard et al. (2009), and following some generic assumption on the physics of the swell propagation in deep water, it was shown that using a back-propagating scheme, it was possible to retrieve the source of the swell system and then generate a propagating field. In this paper, we are proposing a simpler and original approach, by assimilating the SAR data into a given swell field and then using a Kalman Filter/Smoother technique for updating the main parameters of the swell (wavelength, direction, and significant wave height) within the complete field. This method shows very encouraging results which will be confronted with in situ measurements when available.