Despite the fact that the progenitor of fast blue optical transients (FBOTs) is still a subject of debate, FBOTs are sometimes suggested to originate from the core collapse of ultra-stripped stars and be powered by a spinning-down neutron star. Following this consideration, it is expected that the late-time evolution of the progenitor stars can leave important imprints in the circumstellar material (CSM) of FBOTs, due to the strong mass loss of the stars. The interaction of the FBOT ejecta with the CSM can drive a long-lasting shock to generate radio emission, which thus enables us to probe the CSM properties through radio observation although such observations are still rare. Within the framework of the magnetar-powered model, Liu et al. fitted the multi-band optical light curves of 40 FBOTs, and hence, the statistical distributions of the FBOT magnetar and ejecta parameters were obtained. Based on these FBOT population results, we investigate the dependence of the radio emission on the mass-loss rate of the progenitors and evaluate the detectability of radio emission from FBOTs with current and future telescopes. It is found that the distribution of the peak time and peak luminosity of the emission at 8.4 GHz are primarily in the regions of t peak,ν = 102.12±0.63 days and L peak,ν = 1028.73±0.83 erg s−1 Hz−1, respectively. A joint detection of the Zwicky Transient Facility and Very Large Array could achieve success in about 8.7% FBOTs of z ≤ 1. Furthermore, if considering a joint detection of the Chinese Space Station Telescope and the Square Kilometer Array, this rate of success could be increased to about 23.9%.