Abstract: Data fusion is an important way which can be used for accurate inversion of ionosphere. In this paper, a combined ionosphere inversion method by multisource data fusion of space and ground-based observations is proposed. Measurements obtained by ground-based GNSS, low earth orbit(LEO) occultation, beacon and ionosonde are chosen as observation system to do simulation experiment. International Reference Ionosphere (IRI) is chosen to be the background model, while an improved Kriging interpolation method and multiplicative algebraic reconstruction technique (MART) are utilized for effective multisource data assimilation. Inversion results by observation system simulation experiment (OSSE) taking China for example show that it can obtain a good estimation of total electron content (TEC) and ionospheric electron density (IED) by ingesting the multisource data into the IRI model. Errors of the TEC and electron density are significantly reduced after data fusion. Ingestion of ground-based ionosonde and space-based occultation can greatly improves the accuracy of retrieved electron density of the observation system which only relies on the ground-based GNSS. This work will benefit the optimization of the ionospheric observation system in China.