Based on the measured Brillouin gain spectrum, the suitable objective function was chosen. According to a large number of numerically generated Brillouin gain spectra, the influences of signal-to-noise ratio(SNR), scanned frequency interval, number of scanned frequency, frequency scanning range, spectra key parameters g0, vB and ΔvB on the accuracy in the parameters extracted from Brillouin gain spectrum were systematically investigated. The results indicate that the errors in the extracted parameters exponentially decrease with increasing SNR. The errors in the extracted parameters decrease with increasing number of scanned frequency or decreasing scanned frequency interval. The error in the extracted vB increases proportionally with ΔvB. However, ΔvB has no influence on the accuracy in the extracted g0 and ΔvB. If the number of scanned frequency remains constant, then too wide or narrow frequency scanning range will results in significant errors. The optimal frequency scanning range is close to 2ΔvB. This work provides a frame of reference in which you can determine the suitable parameters in Brillouin scattering-based sensing.