The early solar observations of Covington (1947) established a good relation between 10.7 cm solar flux and the presence of sunspots on solar disk. The first spatially resolved observation with a two-element interferometer at arc min resolution by Kundu (1959) found that the radio source at 3 cm has a core-halo structure; the core is highly polarized and corresponds to the umbra of a sunspot with magnetic fields of several hundred gauss, and the halo corresponds to the diffuse penumbra or plage region. The coronal temperature of the core was interpreted as due to gyroresonance opacity produced by acceleration of electrons gyrating in a magnetic field. Since the opacity is produced at resonant layers where the frequency matches harmonics of the gyrofrequency, the radio observation could be utilized to measure the coronal magnetic field. Since this simple interferometric observation, the next step for solar astronomers was to use arc second resolution offered by large arrays at cm wavelengths such as Westerbrock Synthesis Radio Telescope and the Very Large Array, which were primarily built for cosmic radio research. Currently, the Owens Valley Solar Array operating in the range 1-18 GHz and the Nobeyama Radio Heliograph at 17 and 34 GHz are the only solar dedicated radio telescopes. Using these telescopes at multiple wavelengths it is now possible to explore three dimensional structure of sunspot associated radio sources and therefore of coronal magnetic fields. We shall present these measurements at wavelengths ranging from 1.7 cm to 90 cm and associated theoretical developments.