The Internet of Things (IoT) for seismicity in underground mine consists of a combination of sensor network hardware and software, which allows the collection of seismic data and data processing for basic event parameters. A seismic moment tensor is one of the essential parameters in evaluating the seismic hazard. The reliability of the resolved seismic moment tensor depends on numerous factors, and it is important to understand the influence of these disturbances. We focused on influencing factors, including the azimuthal coverage, the accuracy of source coordinate, the mismodeling of velocity structure, and the noise contamination. A set of moment tensor inversions using synthetic data simulated for a double couple source and a superimposed source from virtual IoT was carried out to test the effects of different factors. We concluded that the sensitivity of the moment tensor inversion to the azimuthal coverage is closely related to the type of the source. For the shear event, the inversion does not require a good azimuthal coverage, but for the non-pure shear event, at least 90 • azimuthal coverage is required to get the reliable solution. The full moment tensor is not sensitive to the location accuracy when the source error is less than 60 m, regardless of the type of source. But the double couple deviation increases with the growing error of the location, especially for the fault-slip-related or the shear rupture-dominated event. The erroneous velocity models show nearly no influence on the full moment tensor inversion results. The fit of amplitude spectra does not require a precise alignment of the observed and the synthetic waveforms and is less dependent on the chosen velocity model. The superimposed source is more sensitive to noise than the pure shear source. Although the noise does not necessarily affect the fault plane solution of the event, it is capable to lead an inaccurate seismic moment tensor and mislead the interpretation of the source type. INDEX TERMS Moment tensor inversion, synthetic tests, source model, IoT for seismicity.