Observed steep mass scaling of radio power from the available high mass clusters has ruled out the prospect of detection of 'galaxy groups'. But, the available simulations and observations of thermal emissions show that the groups are merger prone, thus non-virialised, indicating better visibility in the radio waves. Detection of radio emissions from them would help us to understand the scale-dependent particle acceleration mechanisms also groups can be a unique laboratory to test the models of cosmic magnetism and canbe the potential source of WHIMs. So, we have modelled radio emissions from the simulated structures using {\sc{ENZO}}. We present a model for computing magnetic field and for the first time, used the electron energy spectrum from both the Fermi I (DSA) and Fermi II (TRA) mechanisms to compute radio emissions. Computed radio power from more than 200 simulated objects, mass ranging $\geq 10^{13}$ to $2\times 10^{15} M_{\odot}$ show a new mass scaling of $M_{500} \propto P_{1.4\;GHz}^{2.17\pm 0.08}$ and a strong correlation scale of $L_X \propto P_{1.4\;GHz}^{1.08\pm 0.05}$. Both magnetic field and radio power are shown to have adequately replicated the available observations at high mass, allowing us to extend the results to further smaller masses. We report that groups below $10^{14}\;\rm{M_{\odot}}$ show the existence of 10s of nano to a sub-$\mu$G magnetic field and about 10$^{19-23}$ W Hz$^{-1}$ of radio power, much higher than what existing mass scaling predicts. We found that the combined radio power from TRA and DSA electrons can only fit very well to all the observed `radio halos'. Finally, we have implemented this model on a real data set obtained from the Sloan Digital Sky Survey (SDSS). It predicts about 10s to 100s $\mu$Jy/(10$\arcsec$ beam) of radio flux in groups indicating their detectability with existing and aplenty with the future radio telescopes.