Astronomical research dealing with accurate radial velocity measurements need reliable astronomical standards to calibrate the spectrographs and to assess possible systematics. Stellar radial velocity standards offer a reference at the level of a few hundred m/s and are not adequate for most present needs. We aim to show that sunlight reflected by asteroids is a fairly accessible way to record a high-resolution solar spectrum from the whole disk, which can therefore be used as a radial velocity standard and can improve the uncertainties of solar line positions.