In achieving high spectral efficiency, Multiple-Input Multiple-Output (MIMO) performance is strongly influenced by the number of combinations of MxN Tx/Rx antennas and also the MIMO detection algorithm used. Optimization of Rx can be achieved if the order of quadrature phase shift keying (QPSK) implements multi bit-streams and multiplies MxN MIMO levels. Through orthogonal frequency division multiplexing (OFDM), this system will then excel carry all these multi-bit-streams only through the orthogonal subcarrier in parallel transmission. However, multipath and Doppler effect are the main causes frequency shift in the subcarrier of the transmitted OFDM signal, which is called the carrier frequency offset (CFO). Therefore, CFO should be eliminated to increase system performance. This research aims to study modeling estimation techniques for CFO and simulate OFDM synchronization for appropriate frequency sharing 5G standard. The CFO estimation process through OFDM synchronization is carried out to produces zero-mean with a target signal without noise and interference reduction when frequency sharing is implemented, as is the application of using this frequency sharing will often happen on 5G. The research method using three comparison parameters in OFDM-MIMO scenarios, namely Zero-Forcing (ZF), Minimum-Mean-Square-Error (MMSE), and Maximum-Likelihood (ML). The results found are CFO synchronization techniques using the SISO and MIMO operation spatial multiplexing method can well achieve. While the test on the performance of BER Rx-optimum Maximum-Likelihood (ML) versus Eb/No against the use of the number of receiving antennas (MxN) shows that MIMO with more MxN levels indicates lower BER levels small compared to SISO, but both remain stable in synchronization CFO.