In the summer of 2022, the PRime-focus Infrared Microlensing Experiment (PRIME) will start a near infrared (NIR) microlensing survey in the Galactic Center, which cannot be seen by conventional visible observations due to high dust extinction. It will be equipped with a 1.8m telescope (f/2.29) and a 1.45 deg2 (0.5''/pixel) NIR camera and will be installed at the South African Astronomical Observatory (SAAO). Half of the observation time will be used for microlensing survey in the Galactic Center, while the other half will be used for other science, such as the transit surveys for M-dwarfs and the search for counterparts of high-z gamma-ray bursts and gravitational-wave events. A major goal of the PRIME microlensing survey is to reveal the demography of low-mass planets outside the snow-line and the planet frequency in the Galactic Center. PRIME will also help to optimize fields of the NASA Roman microlensing survey and provide insight into the performance of the H4RG-10 detectors that Roman will use. Here, we will present the current results of our estimation of planet yields and the optimal survey strategy by the PRIME microlensing survey. In order to maximize the number of planet detection, we optimize the number of the observation fields and observation cadence, which are in a trade-off relationship. The following steps are used to estimate the detection efficiency of microlensing events and planetary signals: (i) randomly simulate microlensing events toward different directions in the Galactic bulge based on the Galactic model, (ii) generate synthetic data points based on the observation condition and strategy of PRIME, (iii) apply detectability criteria to these simulated events. Then we calculate the statistics for each observation strategy. Finally, we will discuss the optimal survey strategy and predicted planet yields.