Internet of Things(IoT) makes everything intelligently possible. In the IoT era, the application of products such as smart cities and smart homes will provide people with a more convenient and quality life. 5G is the prelude and foundation for achieving this. In the 5G mobile communication system, the base station performs adaptive coding and modulation based on the channel state information fed back by user, and improves the spectrum efficiency by selecting different combinations of modulation types and code rates. Therefore, the channel state information estimated by receiver side needs to be mapped to a value measuring the overall channel state and fed back to the base station. This paper first simulates the modulation and coding schemes(MCS) given in the 5G standard under Gaussian channels, and obtains the SNR-BLER curves of different MCS. Then, the scheme of SNR mapping to CQI is improved, and a new mapping relationship between SNR and CQI is obtained by performing region fitting. Finally, for the adjustment factor optimization process, a new adjustment factor is introduced and the calculation formula for the optimization factor is modified. The BLER is taken as the logarithm and then used Logarithmic results to find the mean square error, thereby reducing mapping errors Compared with the traditional scheme, the proposed scheme has higher throughput, so as to better promote IoT technology.