Particles suspended in a fluid exhibit Brownian motion. This motion is due to collisions of the fluid molecules with the suspended particles. The smaller the particles are, the higher their frequency is. The laser light scattered by particles exhibiting Brownian motion will also fluctuate with time. The fluctuation frequency of the intensity of the scattered light will therefore depend on the size of the particles. Therefore the measured light signal contains information about the particle size. Photon correlation spectroscopy (PCS) is an important method for studying fluctuation frequency of the intensity of the scattered light, in which autocorrelation function estimation plays a key role. Accurate particle size measurements are possible only when the autocorrelation function is of high quality. Based on the character of software correlation, two measures, subtracted average from the signal and adopted non-bias estimation, were taken in this paper to improve the quality of software estimation. The results indicate that it is feasible for getting better autocorrelation estimation using the software autocorrelation method.