Abstract: With the rapid increase in the variety and quantity of electronic devices, man-made radio noise levels have significantly changed. As a result, calculation methods and reference values for man-made radio noise provided by the International Telecommunication Union (ITU) are no longer sufficient to meet the current requirements for the design and performance evaluation of radio systems. Therefore, the measurement and analysis of man-made radio noise in urban areas are important. Software-Defined Radio (SDR) has become a key tool for radio noise measurement. A measurements of urban man-made radio noise based on USRP X310 SDR platform was conducted in Qingdao. An improved interference signal processing method was proposed, a new urban man-made radio noise model was developed, and model accuracy evaluation was performed using the root mean square error (RMSE), mean absolute error (MAE), and goodness of fit (R²). Data analysis reveals that current urban radio noise levels are significantly higher than the reference values reaccommodated by the ITU-R P.372-17. The maximum applicable frequency of the new model has been extended from 250 MHz to 500 MHz. Compared to the ITU-R P.372-17 international standard model, the new model offers superior predictive accuracy and demonstrates high reliability and validity. The newly established model provides foundational support for spectrum planning, interference analysis, and signal detection in urban radio systems. Furthermore, it serves as crucial empirical evidence for the future revision of ITU-R P.372 or the construction of regionalized models.